ISCA - International Speech
Communication Association

ISCApad Archive  »  2014  »  ISCApad #194  »  Events  »  ISCA Events

ISCApad #194

Monday, August 04, 2014 by Chris Wellekens

3-1 ISCA Events
3-1-1(2014-01) INTERSPEECH 2014 Newsletter January 2014


English in Singapore




Attending INTERSPEECH 2014 in Singapore you will probably be glad to know that English is spoken in any corner of the island. Indeed, it is one of the four national languages and the second language spoken in Singapore's homes. According to the last census, in 2010, 89% of the population is literate in English, making of Singapore a very convenient place for tourism, shopping, research or to hold a conference.






Historical context of English in Singapore




The history of Singapore started as the first settlements were established in the 13th century AD [2]. Along the years, Singapore was part of different kingdoms and sultanas until the 19th century, when modern Singapore was founded under the impulsion of the British Empire. In 1819, Sir Thomas Stamford Raffles landed in Singapore and established a treaty with the local rulers to develop a new trading station. From this date, the importance of Singapore continuously grew under the influence of Sir Raffles who, despite not being very present on the island, was the real builder of modern Singapore. Singapore remained under British administration until the Second World War and became a Crown Colony after the end of the conflict. Followed a brief period during which Singapore was part of the Federation of Malaya before becoming independent in 1965 and part of the Commonwealth of Nations.


From this history, Singapore conserved English as one of its four official languages as well as many landmarks that deserve a visit beside of INTERSPEECH. Amongst them, Singapore Botanic Garden, founded in 1859, is internationally renowned [3]. This urban garden of 74 hectares was laid there by Sir Raffles to cultivate and preserve local plants in the tradition of the tropical colonial gardens. Including a Rain Forest, several lakes, an orchid garden and a performance stage, Singapore Botanic Garden is a very popular place to enjoy free concerts on week end afternoons.


Other green spot in the “City in a Garden”, the Padang (field in Malay) was created by Sir Raffles, always him, who planned to reserve the space for public purposes. The place is now famous for the two cricket clubs founded in 1870 and 1883 at both ends of the field and the games that can be watched on weekends.


Amongst the numerous landmarks inherited from the British colonization, the most famous include St Andrew's Anglican cathedral, the Victoria Theater, the Fullerton building, Singapore's City Hall, Old Parliament house, the Central Fire Station and many black and white bungalows built from the 19th century for the rich expatriate families. Some of those bungalows, now transformed in restaurant, will offer you a peaceful atmosphere to enjoy a local diner.






The role of English in Singapore




English has a special place in Singapore as it is the only national language which is not a “mother-tongue”. Indeed, Alsagoff [6] framed English as “cultureless” in that it is “disassociated from Western culture” in the Singaporean context. This cultural voiding makes English an ethnically neutral language used as lingua franca between ethnic groups [5] after replacing the local Malay in this role [4]. Interestingly, English is the only compulsory language of education, and its status in school is that of First Language, as opposed to the Second Language status delegated to the other official languages. By promoting the use of English as working language, the will of the government is to not advantage or disadvantage any ethnic group.


Nevertheless, the theoretical equality stated in the constitution between the four national languages is not always present in practice. For instance, English is overwhelming parliamentary business and some governmental websites are only available in English. Additionally, all legislation is in English only [4].






Singapore English




The standard Singapore English is almost similar to the British English although very cosmopolitan, with 42% of the population born outside the country. Nevertheless, a new standard of pronunciation has been emerging recently [1]. Interestingly, this pronunciation is independent of any external standard and some aspects of it cannot be predicted by reference to British English or any other variety of external English.


The other form of English that you will hear in Singapore is known as Singlish. It is a colorful Creole including words from the many languages spoken in Singapore such as various Chinese dialects (Hokkien, TeoChew, and Cantonese), Malay or Tamil. Many things might be said about Singlish and another newsletter will be especially dedicated to this local variant. Don't miss it!






[1] Deterding, David (2003). 'Emergent patterns in the vowels of Singapore English' National Institute of Education, Singapore. Retrieved 7 June 2013.


(on line January 7
th, 2014)


[3] , (on line January 7th, 2014)


[4] Leimgruber, J. R. (2013). The management of multilingualism in a city-state: Language policy in Singapore. In I. G. Peter Siemund, Multilingualism and Language Contact in Urban Areas: Acquisition development, teaching, communication (pp. 229-258). Amsterdam: John Benjamins.


[5] Harada, Shinichi. 'The Roles of Singapore Standard English and Singlish.' 情報研究 40 (2009): 69-81.


[6] Alsagoff, L. (2007). Singlish: Negotiating culture, capital and identity. In Language, Capital, Culture: Critical studies of language and education in Singapore (pp. 25-46). Rotterdam: Sense Publishers.


3-1-2(2014-02) INTERSPEECH 2014 Newsletter February 2014

At the southern tip of the Malayan Peninsula...



…INTERSPEECH 2014 will be held in Singapore, between Malaysia and Indonesia. In the constitution, Malays are acknowledge as the “indigenous people of Singapore”. Indeed, Malays are the predominant ethnic group inhabiting the Malay Peninsula, Eastern Sumatra, Brunei, coastal Borneo, part of Thailand, the Southern Burmese coast and Singapore. You get now a better understanding of why Malay is one of the four official – and the only national – language of Singapore.



A Malay history of Singapore


It is said that the city of Singapore was founded in 1299 BCE by a Prince from Palembang (South Sumatra, Indonesia), descendant of Alexander the Great. According to the legend, the Prince named the city Singapura (“Lion City”) after sighting a beast on the island. If it is highly doubtful that any Lion ever lived in Singapore outside the zoo, another story tells that the last surviving tiger in Singapore was shot at the bar of the Raffles Hotel in 1902.

Despite this auspicious foundation, the population of Pulau Ujong (the “island at the end”) did not exceed a thousand inhabitants when, in 1819, Sir Thomas Raffles decided to establish a new port to reinforce the British trade between China and India. At this time, the population consisted of different Malay groups (Orang Kallang, Orang Seletar, Orang Gelam, Orang Lauts) and a few Chinese. Nowadays, Malay count for 13.3% of Singapore's population with origins as diverse as Johor, Riau islands (for the Malays Proper), Java (Javanese), Baewan island (Baewanese), Celebes islands (Bugis) or Sumatra (Batak and Minangkabaus).


Malay language


With almost 220 million of speakers, the Malay language in its various forms unites the fifth largest language community in the world [1]. Origins of Malay language can be traced amongst the very first Austronesian languages, back to 2000 BCE [2]. Through the centuries, the major Indian religions brought a number of Sanskrit and Persian words to the Malay vocabulary while islamization of the South East Asia added Arabic influences [3]. Later on, languages from the colonization powers (mainly Dutch and British) and migrants (Chinese and Tamil) contributed to the diversity of Malay influences [4, 5]. In return Malay words have been loaned in other languages, e.g. in English: rice paddy, Orangutan, babirussa, cockatoo, compound, durian, rambutan, etc.

During the golden age of Malay empires, Malay has gained its foothold in territories of modern Malaysia and Indonesia where it became a vector for trade and business. Today, Malay is official language in Malaysia, Indonesia, Brunei and Singapore and is spoken in southern Thailand, Philippines or Cocos and Christmas Islands in Australia [6].

Malay counts a total of 35 phonemes: 6 vowels, 3 diphthongs and 27 consonants [1,5]. As an agglutinative language, its vocabulary can be enriched by adding affixes to the root words [7]. Affixations in Malay consist of prefixation, infixation, suffixation or circumfixation1. Malay languages also have two proclitics, four enclitics and three particles that may be attached to an affixed word [8].

In Singapore, Malaysia, Brunei and Indonesia, Malay is officially written using the Latin alphabet (Rumi) but an Arabic alphabet called Jawi is co-official in Brunei and Malaysia.



Bahasa Melayu in Singapore


Bahasa Melayu (or Malay Language) is one of the four official languages of Singapore, the ceremonial national language and is used in the national anthem or for military commands. However, several creoles remain spoken across the island. Amongst them, Bahasa Melayu Pasar or Bazaar Malay is a creole of Malay and Chinese which used to be the lingua franca and the language for trade between communities [4, 5]. Baba Malay, another variety of Malay Creole influenced by Hokkien and Bazaar Malay is still spoken by around 10,000 people in Singapore.

Today, Bahasa Melayu is the lingua franca among Malay, Javanese, Boyanese, other Indonesian groups and some Arabs in Singapore. It is used as a mean for transmitting familial and religious values amongst the Malay community as well as in “Madrasahs”, mosques and religious schools. However, with 35% of Malay pupils predominantly speaking English at home and a majority of Singaporeans being bilingual in English, Malay is facing competition from English which is taught as first language [5].


Selemat datang ke Singapura / Welcome to Singapore


Opportunities of discovering the Malay culture in Singapore are everywhere. Depending on time and location you might want to taste the Malay's cuisine or one of the succulent Malay cookies while walking around the streets of Kampong Glam or visiting the Malay heritage center.






[1] Tan, Tien-Ping, et al. 'MASS: A Malay language LVCSR corpus resource.' Speech Database and Assessments, 2009 Oriental COCOSDA International Conference on. IEEE, 2009.






[7] B. Ranaivo-Malacon, 'Computational Analysis of Affixed Words in Malay Language,' presented at International Symposium on Malay /Indonesian Linguistics, Penang, 2004 .



1 circumfixation refers here to the simultaneous adding of morphological units, expressing a single meaning or category, at the left and right side of the root word



3-1-3(2014-03) INTERSPEECH 2014 Newsletter March 2014

Tamil and Indian Languages in Singapore



Our fifth step to INTERSPEECH 2014 brings us to the fourth official language of Singapore: Tamil. Today, Indians constitute 9% of the population of Singaporean citizens and permanent residents. They are considered as the third ethnic group in Singapore, although origins of Singaporean-Indians are diverse. Usually locally born, they are second, third, fourth or even fifth generation descendants of Punjabi, Hindi, Sindhi and Gujarati-speaking migrants from the Northern India and Malayalees, Telugu, and Tamil-speaking migrants from the Southern India. This latter group is the core of Singaporean-Indian population with 58% of the Indian community [2, 5].

Before 1819 and Sir Raffles*,

Indianised Kingdoms, such as Srivijaya and Majapahit, radiated over South-East Asia. Influenced by Hindu and Buddhist culture, a large area including Cambodia, Thailand, Malaysia, part of Indonesia and Singapore, formed the Greater India. From this period, Singapore kept some of its most important pre-colonial artifacts such as the Singapore Stone and it is also reported that the hill of Fort Canning was chosen for the first settlement as a reference to the Hindu concept of Mount Meru which was associated to kingship in Indian culture [1].

Under British colony,

Indian migrants arrived to Singapore from different parts of India to fulfill functions such as clerks, soldiers, traders or English teachers. By 1824, 7% of the population was Indian (756 residents). The part of Indian population in Singapore increased until 1860 when it overtook the Malay community and became the second larger ethnic group of 16%. Due to the nature of this migration, Indians in Singapore were predominantly adult men. A settled community, with a more balanced gender and age ratio, only emerged by the mid-20th century [2]. Although the Indian community increased for the following century, its ratio within the Singaporean population decreased until the 1980's, especially when the British withdrew their troupes after Singapore's independence in 1963.

After 1980, the immigration policy aimed at attracting educated people from other Asian countries to settle in Singapore. This change made the Indian population grow from 6.4% to 9%. In addition to this residential population, many ethnic Indian migrant workers temporarily come to work in Singapore (Bangladeshis, Sri Lankans, Malaysian Indians or Indian Indians)[3].

Tamil language

is one of the longest surviving classical languages in the world [8]. Existing for over 2,000 years, Tamil has a rich literature history and was the first Indian language to be declared a classical language by the Government of India in 2004. Earliest records of written Tamil were dated from around the 2nd century BC and, despite the significant amount of grammatical and syntactical change, this language demonstrates grammatical continuity across 2 millennium.

Tamil is the most populous language from the Dravidian language-family, with important groups of speakers in Malaysia, Philippines, Mauritius, South Africa, Indonesia, Thailand, Burma, Reunion and Vietnam. Significant communities can also be found in Canada, England, Fiji, Germany, Netherlands or United States. It is the official language in Indian states of Tamil Nadu, Puducherry, Andaman and Nicobar Islands as well as in Sri Lanka and Singapore.

Like Malay, another local language, Tamil is agglutinative. Affixes are added to words to mark noun class, number, case or verb tense, person, number, mood and voice [7]. Like Finish, not a local language, Tamil sets no limit to the length and extent of agglutination. This leads to long words with a large number of affixes in which its translation might require several sentences in other languages.

Phonology of Tamil is characterized by the use of retroflex consonants and multiple rhotics. Native grammarians classify phonemes into vowels, consonants and a secondary character called āytam. Aytam is an allophone of /r/ or /s/ at the end of an utterance. Vowels are called uyireḻuttu (uyir – life, eḻuttu – letter) and are classified into short (kuṟil), long (neṭil) (with five of each type) and two diphthongs. Unlike most of Indian languages, aspirated and unaspirated consonants are not distinguished in Tamil. Consonants are called meyyeḻuttu (mey—body, eḻuttu—letters) and count three categories: valliṉam—hard, melliṉam—soft or nasal, and iṭayiṉam—medium. Voiced and unvoiced consonants are not distinguished but voice is assigned depending on the position of the consonant in the word.

Tamil writing currently includes twelve vowels, eighteen consonants and one special character for the āytam that combine to form a total of 247 characters.

In Singapore,

Among all the Indian residents in Singapore, 38.8% speaks Tamil daily, 39% speak English, 11% speak Malay, and the remaining 11% speak other Indian languages [2, 4]. Tamil is one the two Indian languages taught as second language (mother tong) in public schools, together with Hindi. It also used in daily newspapers, free-to-air and cable television, radio channels, cinema or theaters [5].

In the multi-cultural environment of Singapore, Tamil influences the other local languages and vice versa. There is especially strong interaction Malay and the colloquial Singaporean English known as Singlish. Singaporean usage of Tamil includes some words from English and Malays while certain words or phrases that are considered archaic in India remain in use in Singapore [2].

During your stay in Singapore,

you can easily get to know Tamil culture through its many aspects. Having a walk in Little India, in which its architecture is protected since 1989, is a great opportunity to be exposed to Tamil music and lifestyle.

The two-storey shop-houses of Singapore's Indian hub host some of the best ambassadors of Indian cuisine. Here you'll find the local version of the Tamil cuisine that has evolved in response to local taste and influences of other cuisines present in Singapore. Other cuisines also include elements of Indian cuisine such as Singapore-Malay cuisine or Peranakan cuisine. Singaporean Tamil must-try include dishes such as achar, curry fish head, rojak, Indian mee goreng, murtabak, roti john, roti prata and teh tarik. Note that other Indian cuisines from Northern India can also be found.















[3] Leow, Bee Geok (2001). Census of Population 2000: Demographic Characteristics. p.47-49.

[4] Singapore Census 2010







* Remember the third step to Singapore



3-1-4(2014-04) INTERSPEECH 2014 Newsletter April 2014

Multilingualism in Singapore


In September (this year), when attending INTERSPEECH, get ready to experience a stimulating multilingual experience. Every year at INTERSPEECH one can hear many languages from all over the world and meet a number of multilingual researchers. However, unlike other editions of this conference, INTERSPEECH 2014 will be held in a highly multilingual environment. From 2000 until 2007, Singapore was ranked the most globalized nation in the world five times1 considering flow of goods, services, people and communications. Indeed, in addition to the four official languages of Singapore, one can also experience many languages across the five continents in the wet markets and shopping malls of Singapore. What’s more interesting is, for many of these Singaporean, they code switch from one language to another naturally and effortlessly.

To speak or not to speak a language

Multilingualism is the ability to use more than one language. When the number of languages is reduced to two, which is the most common form, one talk about bilingualism. There are many ways to use a language so deciding what are the minimum abilities a person should have to be considered as bilingual is a difficult question. For a long time, linguists have limited the definition of bilingual to individuals who had native competency in two languages. This very restrictive definition, which assimilates bilingualism to ambilingualism has now been commonly extended. In its current interpretation, a bilingual person is one who can function at some level in two languages. Whether functioning consists of reading, speaking or, in the case of receptive bilingual, just understanding does not matter. The degree to which the bilingual subject can interact does not matter either and thus, the ability of asking your way in Bahasa Malayu toward a famous conference venue or reading a map in Chinese Mandarin makes you bilingual (as you read these lines).


Bilingualism and Multilingualism

Amongst the most famous multilingual speakers, is Giuseppe Mezzofanti, a 19th century Italian Cardinal, was reputed to speak 72 languages. If you consider claiming to know 72 languages to be a bit too gimmicky, consider this other case of a Hungarian interpreter during the cold war who was able to speak 16 languages (including Chinese and Russian) by the age of 86 [1]. Nevertheless, not all multilingual are hyper-polyglot as being able to learn 12 languages or more is not so common.

The complex mechanism of learning a new language is not clear yet and many questions remain, regarding a possible age limitation or the relationship between already mastered languages and the ease of learning a new one. Nevertheless, before considering learning an additional language you should be aware that this is a complicated process that has many effects. It might of course open your mind to other cultures and ways of thinking, but more importantly, it can deeply modify your brain. Neuroscience is a very active field when it comes to multilingualism. The powerful imaging tools available as well as the observation of subjects affected by trauma have led to a better understanding of the language learning process. Different language areas have been located within the brain and an augmented plasticity of the over-whole structure has been demonstrated for the case of multilingual speakers [2, 3]. Interestingly, the brain structure of simultaneous bilinguals, who learned two languages without formal education during childhood, is similar to that of monolingual subjects. On the contrary, learning a second language after gaining proficiency in the first language modifies the brain structure in an age-dependent manner [4].

Amongst the benefits of multilingualism, it has been shown that it increases the ability to detect grammatical errors and improve the reading ability. In the case of bimodal subjects, who use both a spoken language and a signed language, bilingualism improves the brain's executive function that directs the attention processes used for planning, solving problems and performing various mentally demanding tasks [2].

When it comes to society, multilingualism is the fact of having several languages spoken in a reduced area. Speakers don't have to interact or to be multilingual themselves. This phenomenon is observed in many countries or cities in the world and can take different forms. When a structural distribution of languages exists in the society, one talk about polyglossia. Multipart-lingualism refers to the case where most speakers are monolingual and speak different languages, while omnilingualism, the less common, describes the situation where no structural distribution can be observed and that it is nearly impossible to predict which language is going to be spoken in a certain context. That's the former one that you are going to experience in Singapore.


Multilingualism in Singapore

The city-state of Singapore is born from multi-multiculturalism and multilingualism. No wonder then that the choice of having four official languages was thought to be a central piece of the community harmony. In 1966, considering the lack of natural resources and the dominance of international trade in their local economy, Singaporean leaders decided to reinforce English as a medium of economic development [5]. In 1987, English was officially acknowledge as first language while others official languages were referred to as mother tongues [6]. Singapore's bilingualism is thus described as “English-knowing” because of the central role of English [7].

The success of Singapore is said to be partly the result of the language policy which fueled the globalization process of the Lion City [8]. Indeed, the promotion of English as the common neutral language amongst ethnic groups in Singapore facilitated Singapore's integration into the world economy. On the other hand, predominance of English has raised concerns about the decreasing usage of mother tongues and the demise of traditional cultural values [8].

In the last 30 years, language education has been undertaken by the state as one way to control globalization and to reduce the impact of Western culture that tends to replace Asian culture [8]. The growing importance of Western culture in Singapore is reflected by the shift in home languages towards English. Therefore, to reinforce the Asian cultural identity, Singapore's government has emphasized the learning of mother tongues. This policy is considered controversial to some as it led to the popularity of Mandarin Chinese and Bahasa Malayu at the expense of the loss of many other Chinese and Malay language varieties. It is no doubt a delicate and challenging trade-off between preserving language diversity and enforcing common languages for the convenience of communication and economic development.


Moving away from a language policy stemming from boosting economic development will probably take time. The implicit role of languages in Singapore’s multi-ethnic society can be significant yet complex. However, there is no doubt that Singaporeans consider multilingualism as a major component of their national identity that relies not only on the four official languages. One way to realize that during your stay with us is to ask any Singaporean about her/his language background and to get immersed in the rich diversity of spoken languages in Singapore .



[1] (accessed on 7 April, 2014)

[2] (accessed on 7 April, 2014)

[3] (accessed on 7 April, 2014)

[4] Klein, D., Mok, K., Chen, J. K., & Watkins, K. E. (2013). Age of language learning shapes brain structure: A cortical thickness study of bilingual and monolingual individuals. Brain and language.

[5] 'Interview: Chinese Language education in Singapore faces new opportunities'. People's Daily Online. 2005-05-13. (accessed on 7 April, 2014)

[6] Pakir, A. (2001). Bilingual education with English as an official language: Sociocultural implications. GEORGETOWN UNIVERSITY ROUND TABLE ON LANGUAGES AND LINGUISTICS 1999, 341.

[7] Tupas, R. (2011). English knowing bilingualism in Singapore: Economic pragmatics, ethnic relations and class. English language education across greater China, 46-69.

[8] (accessed on 7 April, 2014)

1 A.T. Kearney/Foreign policy globalization index accessed April 3, 2014



3-1-5(2014-05) INTERSPEECH 2014 Newsletter May 2014

Language Education





“If you talk to a man in a language he understands, that goes to his head. If you talk to him in his own language, that goes to his heart.”


Nelson Mandela



There is no doubt that this quote will continue to inspire generations of language learners despite the recent advancement of statistical machine translation. Before learning more about the latest discovery in language learning and machine translation during INTERSPEECH 2014, the newsletter of this month is dedicated to language education in Singapore.



Brief history of language education


Discovery of the “New World” is sometimes considered as the starting point of globalization. Of course travelers, merchants and scholars didn't wait that long to study languages but, surprisingly, theorization of language education is quite recent. In the 17th century, Latin was commonly used for education, language and religion in the Western countries. Its teaching was almost exclusively done through grammatical aspects until Jan Amos Komenský, a Czech teacher and educator, created a complete course for learning this language [12]. Jan Amos Komenský major contributions also include the invention of the primer and textbook which are now widely used to teach reading and languages.

During the 19th and 20th centuries, research on language education sped up and led to a large number teaching practices which were supposed to improve the experience of language learners. In 1963, Anthony [14] proposed a three-layer hierarchical framework to describe language teaching including approaches, methods and techniques. Approaches are related to general concepts about the nature of languages, while methods refer to the over-whole plan of the language teaching organization which is implemented in class through techniques which aim at achieving short term objectives. Anthony's framework was later extended by Richards and Rogers [15], who especially extended the concepts of methods and techniques to designs and procedures that were intended to be more specific and less descriptive.

Amongst the most popular, the structural methods consider languages through the prism of grammar, functional methods focus more on languages as a vehicle to accomplish certain functions and interactive ones emphasize on social relations such as acts, negotiation and interactions. Of course this list is not exhaustive and do not address the complexity of the whole range of existing methods.



Language education in the world


In Africa, where most countries used to be colonized, language policies strongly depend on the former colonial power and its tolerance of the local languages [16] but also on the post-independence political evolution, on the socio-linguistic contour of each country and on the level of education. During colonization, the French, Portuguese and Spanish used to teach their language at all levels and from the first day of school. The Germans did promote their language while giving prominence to local languages in the first years of schooling and the British conducted the first year of education in the local language before changing it to English in the following years. In some parts of Western Africa, British even encourage the teaching of certain languages such as Hausa, Igbo, Yoruba, Efik, Ga, or Ewe but kept English as a reference point. After independence, most of the African countries considered reforming education to promote indigenous languages but in a lot of cases, those politics have been questioned as teaching in the mother tongue could weaken the national unity. Of course, it has been easier for the few monolingual countries (Somalia, Burundi, Rwanda, Botswana, Lesotho, Swaziland, Madagascar, etc.) to promote education in their native languages while some multilingual countries have chosen to develop regional languages. For instance, Zambia uses six zonal languages in education, Zaire four, and Togo two. For economical reasons, English and French are still taught across the former colonies and are still strong factor of regional cohesion.


Bilingual education in South America mostly refers to the teaching of a “mainstream” language such as Spanish or Portuguese to non-Spanish and non-Portuguese speaking people [17]. It usually follows a “transition” or “maintenance” model. In the first one, the official language progressively replaces the mother tongue while the second one makes use of the two languages through the whole curriculum. Language distribution in South America is mainly characterized by the fact that many populations are located in isolated area where communication and outside contact are poor and thus, monolingualism is prevalent. In this context, a number of countries have launched bilingual education programs in the 1970's. Those programs have provided good results since bilingual education improved education amongst indigenous children.


North America’s modern history includes periods of colonialism too. However, the language education evolved in a very different way and strongly differ between Canada and U.S.. In the 19th century, U.S. was especially friendly towards bilingualism as immigrant communities commonly maintained and published in their native language [19]. Starting from the 1880's, and due to a huge influx of non-English speaking immigrants, English was used to develop an “American” identity. Monolingualism as then become the norm and second language learning is still uncommon before high school. Therefore, only 15 to 20 percent of Americans consider themselves bilingual compared to 56 percent of European (European commission survey, 2006). The most common second languages taught in the U.S. include Spanish, due to the large number of recent Spanish-speaking immigrants, followed by French, German, Latin, Mandarin, Italian and Japanese, in descending order of frequency. As a multilingual country, Canada allows two languages of instruction: English by default and French in the case of “Francophone children whose parents qualify for minority language rights” [18]. Additionally, aboriginal languages can be taught as a second language as all students are required to learn a second language from 9 to 14 years old.


In Europe, all children studied at least one foreign language as part of their compulsory curriculum except in Ireland where instruction includes English and Irish, both considered a native language, and a third European language. In all European countries, English is by far the most commonly learned language before French, Spanish, German and Russian. On average, children start learning a second language between 6 and 9 years old [20]. This age has strongly decreased in the last 15 years and it is now common for children to start learning in pre-school. However, the weekly number of hours spent learning a language did not really increased in the same time. Due to the multilingual context, and in order to encourage cross-border exchanges, the European Union strongly encourages learning of foreign languages and, on average, in 2009/10, 60.8% of lower secondary education students were learning two or more foreign languages. From a local point of view, almost all European countries have regional languages and more than half of the countries use partial immersion to teach both the minority and the state language.


In South-East Asia, 550 millions inhabitants speak hundreds of languages including local (Javanese, Hmong for example), national (Khmer, Thai, for instance) and regional languages (varieties of Chinese and Malay) [2]. Amongst the eleven countries of this region, all except Thailand have endured colonization and been exposed to European languages: Dutch in Indonesia, English in Brunei, Burma, Malaysia, Philippines and Singapore, French in Cambodia, Laos and Vietnam and Portuguese in East Timor. After decolonization, most governments used languages to strengthen national cohesion and forge a national identity. All eleven Southeast Asian countries have included English in education, often as a foreign language. In certain countries, however, instruction is given in the national language while sciences and technologies are taught in a foreign language, for instance English in Myanmar and French in Laos.





Under British colonial rule, school systems in the main four languages, namely Chinese Mandarin, English, Malay and Tamil, cohabited in Singapore [4]. After World War II, the schools were gradually brought under the control of the government, which decided to establish one of the existing languages as lingua franca to strengthen the national unity. Amongst the possible languages, Malay was considered a good choice given the integration of Singapore to the Federation of Malaya, and Hokkien was already spoken by the majoritys of Chinese Singaporeans. However, the government decided to chose English as it was both a tool for economic development and an ethnic neutral language in the context of Singapore’s multi-ethnic population including Chinese, Malay and Indian.

The bilingual education policy was officially introduced in 1966 with the possibility to teach English as a first or second language. However, schools teaching English as a second language declined rapidly as English was considered a key element for professional success. By 1986, there remained a single class of 28 secondary school students following a curriculum in Malay and Malay-medium schools, came to a natural demise like the Tamil-medium schools in 1982. Chinese-medium schools were removed by the government [4]. The government then officially defined English as the first language and the three other official languages as mother tongues. In a will of preserving the Asian culture in Singapore, the government imposed the learning of the mother tongue as second language. This mother tongue is determined for each student depending on her/his ethnicity. Therefore, Malay Singaporean have to learn Bahasa Malayu, Chinese learn Mandarin while Indian from a Dravidian language learn Tamil. Indian is a special case as non-Vernacular Languages like Hindi, Bengali, Punjabi, Gujarati and Urdu can be chosen as a mother tongue by non-Tamil but the state does not provide teachers in those languages [3]. On the opposite, all Singaporean Chinese have to learn Mandarin despite the various linguistic backgrounds present in the local community. Due to this policy, importance of non-Mandarin Chinese languages strongly decreased in the last 50 years and Mandarin is now the first-spoken language in Singaporean homes. Since 2002, Chinese associations in Singapore propose dialect classes in order to reconnect the population with its Chinese culture and enable the younger generation to talk to elderly [3]. A third language can be learn starting from secondary school for which students can chose amongst Mandarin (for non-Chinese), Malay (for non-Malays), Bahasa Indonesia (for non-Malays), Arabic, Japanese (only for Chinese), French, German and Spanish [4,6].

Although it is one of the reasons of Singapore's exceptional economic success, bilingual policy has been, according to the government itself, a cultural failure. By promoting English as a business and inter-ethnic language, the bilingual policy made other languages less attractive to the younger generation. Additionally, the mother tongues have been taught as discipline while using methods developed for a native language. As a consequence, many Singaporean students don't see the point of learning a language which is not a vector of culture but only a subject of study. Realizing this mistake, the government recently decided to make language learning more interesting and IT-based. For example, language learning through the use of smart phones and on-line computer games [5,10]

From a wider perspective, Singapore is unique in Asia as it has a strong national education system at a moment where other countries massively privatize instruction [11] and also because of the way, probably unparalleled in any other developed country, the state’s intervention changed the people’s language and speech patterns [1].




[1] Language, Society and Education in Singapore: Issues and Trends (Second Edition); S. Gopinathan, Anne Pakir, Ho Wah Kam and Vanithamani Saravanan (Eds.); Times Academic Press, Singapore, 1998

[2] Language Education Policies in Southeast Asia, T Clayton, elsevier









[11] Globalization and Multilingualism in Singapore: Implications for a hybrid identity



[14] Anthony, E. M. (1963). 'Approach, Method, and Technique'. ELT Journal (2): 63–43. doi:10.1093/elt/XVII.2.63

[15] Richards, Jack; Rogers, Theodore (2001). Approaches and Methods in Language Teaching. Cambridge: Cambridge University Press. ISBN 978-0-521-00843-3









3-1-6(2014-06) INTERSPEECH 2014 Newsletter June 2014

Kristang, an endangered Portuguese creole



Amongst the many languages spoken in Singapore, Kristang is probably one of the less likely you might hear when attending INTERSPEECH 2014. Indeed, this Portuguese creole originating from Malacca, Malaysia, is spoken by less than five hundreds persons and categorized as an endangered language according to the UNESCO Atlas of the World’s Languages in Danger.


Some background about creoles


Creole languages can easily be mixed up with pidgins. The fundamental difference between those two categories of languages stands in the existence of native speakers [9]. A pidgin is a language without native speakers that is developed in areas where several languages co-exist, while a creole is taught by a generation to another and thus is native language for a community of speakers. Therefore, pidgins emerge in societies where several mother tongues are used in the same place to facilitate mutual understanding and can be referred to as “contact languages”. When spoken by many people it might become a creole if transmitted over generations. In some cases, creoles can even replace the existing mix of languages to become the official language such as Krio in Sierra Leone and Tok Pisin in Papua New Guinea.

From the linguistic perspective, both pidgins and creoles are derived from several languages and generally involve simplification of the vocabulary and syntax (grammar). They also often include considerable phonological variations and fulfill fewer functions than the original languages.

Before the 1930's, pidgins and creole were mostly ignored by linguists. Recently, more attention has been paid to these languages. In 1997, Hancock [9] listed 127 pidgins and creoles languages: thirty-five described as English-based, fifteen French-based, fourteen Portuguese-based, seven Spanish-based, six German-based, five which are based on Dutch language, three on Italian and the rest based on a variety of languages such as Russian, Chinese or Malay. Most of the creoles and pidgins are distributed in the equatorial belt where contact between languages is facilitated by oceans and trade.





Kristang is a Portuguese-based creole language influenced by Malay, English and other languages spoken on the Malaya peninsula [1]. Called Papia Kristang or Christao, this creole originated from Malacca in 1511, when the Portuguese explorer, Alfonso de Albuquerque, conquered the city [7]. Strategically located on the spice trade routes of South-East Asia, Malacca was a way for Portuguese to challenge the dominance of Venice in the trading or rare spices [8]. In order to ensure the loyalty of the local population and to provide manpower, Alfonso de Albuquerque encouraged marriages between Portuguese men and Malay women [2]. In 1641, Portuguese lost Malacca to the Dutch and Dutch men married local “Portuguese” women and embraced their Catholic faith. The mix of Malay, Portuguese and Dutch were known as “Malacca-Portuguese” or Jenti Kristang (Kristang people) speech community [8]. After the Dutch captured the city, the Kristang community not only preserved the language but also, through migrations, influenced other languages such as Macanese, the creole language spoken in Macao, another Dutch colony.


Although Kristang has no written form and has never been taught in school, it has been passed down from generation to generation, through daily usage and by being used in church services [2]. A first proposal for standard orthography was made in late 1980's by Alan Baxter in which he suggested to use Malay orthography. In the 1990s, Joan Marbeck's book 'Ungua Andanza' was published, with a “Luso-Malay” orthography. The grammatical structure of Kristang is very close to Malay but a large part of the vocabulary (~95%) is Portuguese, so Kristang is generally quite recognizable to speakers of European Portuguese although many words are considered archaic. Perhaps because of cultural exchanges along trade routes, Kristang has a lot of similarities with other Portuguese-based creoles spoken in Indonesia and East-Timor. According to Baxter [8], Kristang's pronunciation is very close to the colloquial Malacca Malay, for instance, the vowel /e/ is usually pronounced as an /i/ when followed by a syllable with a /i/, for example, penitensia ('penitence') is pronounced [piniˈteɲsia].


Nowadays, Kristang counts 5,000 speakers in Malacca and 400 in Singapore; it is also spoken in some parts of Indonesia and in Australia (region of Perth) due to migrations. Kristang is considered the “last vital variety of a group of East and Southeast Asian Creole Portuguese languages” [6] and categorized as one of the endangered languages in Malaysia.

In order to revitalize the language, publications of dictionaries, phrase-books, and language documentation efforts are encouraged. Social media are also used as a way to promote the use of Kristang with Facebook pages such as “Keep Kristang alize” and “Yo Falah Linggu Kristang” (I speak the Kristang language).


Although Kristang is only spoken by a handful of people in Singapore these days, and most people (including local Singaporeans and the Portuguese and the Dutch) are virtually unaware of its existence, Kristang symbolizes the rich multilingualism rooted in Singapore in a historical context.









[6] Stefanie Pillai, Wen-Yi Soh, Angela S. Kajita , “Family language policy and heritage language maintenance of Malacca Portuguese Creole,” in Language and Communication, 2014, in press

[7] Bryan W. Husted, “Globalization and cultural change in international business research,” in Journal of International Management, 2003, pp. 427-433

[8] Ei Leen Lee, “Language maintenance and competing priorities at the Portuguese Settlement Malacca,” in Ritsumeikan Journal of Asia Pacific Studies, 2011, vol. 30, pp. 77-99

[9] Syarfuni Syarfuni, “Pidgins and creoles languages,” in Visipena, vol. 2, issue 1, 2011


3-1-7(2014-07) INTERSPEECH 2014 Newsletter July 2014

Singlish: a living example of multilingualism blending the East and West


Singlish (colloquial Singapore English) is a vivid and colorful creole example of how languages and speakers interact and mingle: while Singlish is a language variety of English, it is interleaved with slangs from languages such as Hokkien, Teochew, Cantonese, Malay and Tamil, and heavily influenced by Chinese grammar, phonology, and prosody. The complexity of Singlish exemplifies the challenges speech researchers face in developing spoken language technologies to automatically identify, transcribe, and parse colloquial and conversational speech.


Singlish is semi-tonal, as all words of Chinese origin retain their original words, while original English words as well as Malay and Tamil words are non-tonal. In addition, although most varieties of English are stressed-time, Singlish is syllable-timed, giving Singlish a rather staccato feel.


Singlish phonology is primarily British based, with influence of Chinese phonology. For example, the dental fricatives /θ/ and /ð/ are sometimes merged with /t/ and /d/ in certain contexts, so that three sounds like tree and then sounds like den [16]. The voiceless stops /p/, /t/, /k/ are also sometimes unaspirated as in Chinese languages [18]. There is generally no distinction between the non-close front monophthongs, so pet and pat are pronounced the same /pɛt/ [17].


At the vocabulary level, there is often inter-mixing of multiple languages. For example, damn shiok is a slang blending English and Punjabi, used to express extreme pleasure or satisfaction, often in the context of food. Mixing of languages is also reflected from location names. For example, Toa Payoh (literal translation: big swamp), a central district in Singapore, mixes Hokkien and Malay (toa is big in Hokkien and payoh is swamp in Malay). Reduplication is also used in Singlish, which is influenced by Chinese and Malay. Adjectives of one or two syllables can also be repeated for intensification. For example, “You go take the small-small one ah.” (Retrieve the smaller item, please.) The frequent use of already (pronounced more like oreddy) in Singapore English is probably a direct influence of the Hokkien liao particle [18]. For example, “Aiyah, cannot wait any more, must go oreddy.” (Oh dear, I cannot wait any longer. I must leave immediately.)


Singlish is topic-prominent like Chinese and Japanese, meaning that Singlish sentences often begin with a topic followed by a comment of new information. For example, Dis country weather very hot one.” (In this country, the weather is very warm.) The topic can be omitted when the context is clear, resulting in constructions that appear to be missing a subject. For example, “No good lah” (This isn’t good.)


Singlish is also known for its colorful usage of interjections from Chinese and Malay influence (examples in previous sentences examples include ah, aiyah, lah). ”lah” is probably the most famous one and a stereotypical interjection which appears to be ubiquitous to non-native speakers of Singlish. It may originate from the Hokkien character (), though its usage in Singapore is also influenced by its occurrence in Malay [19]. “lah” has many different usages. It is often used to soften ones tone. For example, “Cannot lah”, “Just drink lah”. It can also be used to indicate impatience with a low tone; e.g., “Eh, hurry up lah!” It can also be used for reassurance: Okay lah. (It's all right. Don't worry about it.) Yet, it can also be used to curse people. For example, “Go and die lah”.


Although Singlish is typically not used in official settings (e.g. school lectures and mainstream media generally use Standard Singapore English), Singlish is quite prevalent in day-to-day interactions with peers, siblings, parents, and elders. It is an effective means to establish rapport (for example, during military service) or for humorous effects for TV and radio shows. From a linguistic perspective, Singlish is a living example of multilingualism in Singapore blending the East and the West.


3-1-8(2014-09-14) CfP INTERSPEECH 2014 Singapore URGENT Action Required

Interspeech 2014


September 14-18, 2014


INTERSPEECH 2014 paper submission deadline is on 24 March 2014.  There will be no extension of deadline.  Get ready your paper submissions and gear up for INTERSPEECH in Singapore.



INTERSPEECH is the world's largest and most comprehensive conference on issues surrounding the science and technology of spoken language processing, both in humans and in machines.

The theme of INTERSPEECH 2014 is 'Celebrating the Diversity of Spoken Languages'. INTERSPEECH 2014 emphasizes an interdisciplinary approach covering all aspects of speech science and technology spanning basic theories to applications. In addition to regular oral and poster sessions, the conference will also feature plenary talks by internationally renowned experts, tutorials, special sessions, show & tell sessions, and exhibits. A number of satellite events will take place immediately before and after the conference. Please follow the details of these and other news at the INTERSPEECH website

We invite you to submit original papers in any related area, including but not limited to:

1: Speech Perception and Production

2: Prosody, Phonetics, Phonology, and Para-/Non- Linguistic Information

3: Analysis of Speech and Audio Signals

4: Speech Coding and Enhancement

5: Speaker and Language Identification

6: Speech Synthesis and Spoken Language Generation

7: Speech Recognition - Signal Processing, Acoustic Modeling, Robustness, and Adaptation

8: Speech Recognition - Architecture, Search & Linguistic Components

9: LVCSR and Its Applications, Technologies and Systems for New Applications

10: Spoken Language Processing - Dialogue, Summarization, Understanding

11: Spoken Language Processing -Translation, Info Retrieval

12: Spoken Language Evaluation, Standardization and Resources 

A detailed description of these areas is accessible at:


Paper Submission

Papers for the INTERSPEECH 2014 proceedings should be up to 4 pages of text, plus one page (maximum) for references only. Paper submissions must conform to the format defined in the paper preparation guidelines and provided in the Authors’ kit, on the INTERSPEECH 2014 website, along with the Call for Papers. Optionally, authors may submit additional files, such as multimedia files, which will be included in the official conference proceedings USB drive. Authors must declare that their contributions are original and are not being submitted for publication elsewhere (e.g. another conference, workshop, or journal). Papers must be submitted via the online paper submission system, which will be opened in February 2014. The conference will be conducted in English. Information on the paper submission procedure is available at:

There will be NO extension to the full paper submission deadline.


Important Dates

Full Paper submission deadline


24 March 2014

Notification of acceptance/rejection


10 June 2014

Camera-ready paper due


20 June 2014

Early registration deadline


10 July 2014

Conference dates


14-18 Sept 2014

We look forward to welcoming you to INTERSPEECH 2014 in Singapore!


Helen Meng and Bin Ma

Technical Program Chairs





Email:— For general enquiries


Conference website:


3-1-9(2014-09-14) CfP Speech Technology for the Interspeech App

Call for Proposals

Speech Technology for the Interspeech App

During the past Interspeech conference in Lyon, a mobile application (app) was provided for accessing the conference program, designing personal schedules, inspecting abstracts, full papers and the list of authors, navigating through the conference center, or recommending papers to colleagues. This app was designed by students and researchers of the Quality and Usability Lab, TU Berlin, and will be made available to ISCA and to future conference and workshop organizers free-of-charge. It will also be used for the upcoming Interspeech 2014 in Singapore, and is available under both iOS and Android.

In its current state, the app is limited to mostly touch-based input and graphical output. However, we would like to develop the app into a useful tool for the spoken language community at large, which should include speech input and output capabilities, and potentially full spoken-language and multimodal interaction. The app could also be used for collecting speech data under realistic environmental conditions, for distributing multimedia examples or surveys during the conference, or for other research purposes. In addition, the data which is being collected with the app (mostly interaction usage patterns) could be analyzed further.

The Quality and Usability Lab of TU Berlin would like to invite interested parties to contribute to this development. Contributions could be made by providing ready-built modules (e.g. ASR, TTS, or alike) for integration into the app, by proposing new functionalities which would be of interest to a significant part of the community, and preferably by offering workforce for such future developments.

If you are interested in contributing to this, please send an email with your proposals to

by October 31, 2013. In case that a sufficient number of interested parties can be found, we plan to submit a proposal for a special session around speech technology in mobile applications for the upcoming Interspeech in Singapore.

More information on the current version of the app can be found in: Schleicher, R., Westermann, T., Li, J., Lawitschka, M., Mateev, B., Reichmuth, R., Möller, S. (2013). Design of a Mobile App for Interspeech Conferences: Towards an Open Tool for the Spoken Language Community, in: Proc. 14th Ann. Conf. of the Int. Speech Comm. Assoc. (Interspeech 2013), Aug. 25-29, Lyon.


3-1-10(2014-09-14) INTERSPEECH 2014 Tutorials
--- September 14-18, 2014

The INTERSPEECH 2014 Organising Committee is pleased to announce 
the following 8 tutorials presented by distinguished speakers 
at the conference and will be offered on Sunday, 14 September 2014.
All Tutorials will be of three (3) hours duration and require 
an additional registration fee (separate from the conference registration fee). 

    • Non-speech acoustic event detection and classification
    • Contribution of MRI to Exploring and Modeling Speech Production
    • Computational Models for Audiovisual Emotion Perception
    • The Art and Science of Speech Feature Engineering
    • Recent Advances in Speaker Diarization
    • Multimodal Speech Recognition with the AusTalk 3D Audio-Visual Corpus
    • Semantic Web and Linked Big Data Resources for Spoken Language Processing
    • Speech and Audio for Multimedia Semantics


Additionally, the ISCSLP 2014 Organising Committee welcomes 
the INTERSPEECH 2014 delegates to join the 4 ISCSLP tutorials 
which will be offered on Saturday, 13 September 2014.

    • Adaptation Techniques for Statistical Speech Recognition
    • Emotion and Mental State Recognition: Features, Models, System Applications and Beyond
    • Unsupervised Speech and Language Processing via Topic Models
    • Deep Learning for Speech Generation and Synthesis

More information available at:

Tutorials Description

T1: Non-speech acoustic event detection and classification

    The research in audio signal processing has been dominated by speech research, 
    but most of the sounds in our real-life environments are actually non-speech 
    events such as cars passing by, wind, warning beeps, and animal sounds. 
    These acoustic events contain much information about the environment and physical
    events that take place in it, enabling novel application areas such as safety,
    health monitoring and investigation of biodiversity. But while recent years 
    have seen wide-spread adoption of applications such as speech recognition and 
    song recognition, generic computer audition is still in its infancy.

    Non-speech acoustic events have several fundamental differences to speech, 
    but many of the core algorithms used by speech researchers can be leveraged 
    for generic audio analysis. The tutorial is a comprehensive review of the field 
    of acoustic event detection as it currently stands. The goal of the tutorial is 
    foster interest in the community, highlight the challenges and opportunities 
    and provide a starting point for new researchers. We will discuss what acoustic 
    event detection entails, the commonalities differences with speech processing, 
    such as the large variation in sounds and the possible overlap with other sounds. 
    We will then discuss basic experimental and algorithm design, including descriptions 
    of available databases and machine learning methods. We will then discuss more 
    advanced topics such as methods to deal with temporally overlapping sounds and 
    modelling the relations between sounds. We will finish with a discussion of 
    avenues for future research.

    Organizers: Tuomas Virtanen and Jort F. Gemmeke

T2: Contribution of MRI to Exploring and Modeling Speech Production

    Magnetic resonance imaging (MRI) provides us a magic vision to look into 
    the human body in various ways not only with static imaging but also with 
    motion imaging. MRI has been a powerful technique for speech research to 
    study finer anatomy of the speech organs or to visualize true vocal tracts 
    in three dimensions. Inherent problems of slow image acquisition for speech 
    tasks or insufficient signal-to-noise ratio for microscopic observation have 
    been the cost for researchers to search for task-specific imaging techniques. 
    The recent advances of the 3-Tesla technology suggest more practical solutions 
    to broader applications of MRI by overcoming previous technical limitations. 
    In this joint tutorial in two parts, we summarize our previous effort to accumulate 
    scientific knowledge with MRI and to advance speech modeling studies for future 
    development. Part I, given by Kiyoshi Honda, introduces how to visualize the 
    speech organs and vocal tracts by presenting techniques and data for finer static 
    imaging, synchronized motion imaging, surface marker tracking, real-time imaging, 
    and vocal-tract mechanical modeling. Part 2, presented by Jianwu Dang, focuses on 
    applications of MRI for phonetics of Mandarin vowels, acoustics of the vocal tracts 
    with side branches, analysis and simulation in search of talker characteristics, 
    physiological modeling of the articulatory system, and motor control paradigm 
    for speech articulation.

    Organizers: Kiyoshi HONDA and Jianwu DANG

T3: Computational Models for Audiovisual Emotion Perception

    In this tutorial we will explore engineering approaches to understanding human 
    emotion perception, focusing both on modeling and application. We will highlight 
    both current and historical trends in emotion perception modeling, focusing on 
    both psychological and engineering-driven theories of perception 
    (statistical analyses, data-driven computational modeling, and implicit sensing). 
    The importance of this topic can be appreciated from both an engineering viewpoint, 
    any system that either models human behavior or interacts with human partners must 
    understand emotion perception as it fundamentally underlies and modulates our 
    communication, or from a psychological perspective, emotion perception is also used 
    in the diagnosis of many mental health conditions and is tracked in therapeutic 
    interventions. Research in emotion perception seeks to identify models that describe 
    the felt sense of ‘typical’ emotion expression – i.e., an observer/evaluator’s attribution 
    of the emotional state of the speaker. This felt sense is a function of the methods through 
    which individuals integrate the presented multimodal emotional information. 
    We will cover psychological theories of emotion, engineering models of emotion, 
    and experimental approaches to measure emotion. We will demonstrate how these modeling 
    strategies can be used as a component of emotion classification frameworks and how 
    they can be used to inform the design of emotional behaviors.

    Organizers: Emily Mower Provost and Carlos Busso

T4: The Art and Science of Speech Feature Engineering

    With significant advances in mobile technology and audio sensing devices, 
    there is a fundamental need to describe vast amounts of audio data in terms 
    of well representative lower dimensional descriptors for efficient automatic 
    processing. The extraction of these signal representations, also called features, 
    constitutes the first step in processing a speech signal. The art and science of 
    feature engineering relates to addressing the two inherent challenges - extracting 
    sufficient information from the speech signal for the task at hand and suppressing 
    the unwanted redundancies for computational efficiency and robustness. 
    The area of speech feature extraction combines a wide variety of disciplines like 
    signal processing, machine learning, psychophysics, information theory, linguistics and physiology. 
    It has a rich history spanning more than five decades and has seen tremendous advances 
    in the last few years. This has propelled the transition of the speech technology from 
    controlled environments to millions of end user applications.

    In this tutorial, we review the evolution of speech feature processing methods, 
    summarize the recent advances of the last two decades and provide insights into the 
    future of feature engineering. This will include the discussions on the spectral 
    representation methods developed in the past, human auditory motivated techniques 
    for robust speech processing, data driven unsupervised features like ivectors and 
    recent advances in deep neural network based techniques. With experimental results, 
    we will also illustrate the impact of these features for various state-of-the-art 
    speech processing systems. The future of speech signal processing will need to address 
    various robustness issues in complex acoustic environments while being able 
    to derive useful information from big data.

    Organizers: Sriram Ganapathy and Samuel Thomas

T5: Recent Advances in Speaker Diarization

    The tutorial will start with an introduction to speaker diarization giving a general 
    overview of the subject. Afterwards, we will cover the basic background including 
    feature extraction, and common modeling techniques such as GMMs and HMMs. 
    Then, we will discuss the first processing step usually done in speaker diarization 
    which is voice activity detection. We will consequently describe the classic approaches 
    for speaker diarization which are widely used today. We will then introduce state-of-the-art 
    techniques in speaker recognition required to understand modern speaker diarization techniques. 
    Following, we will describe approaches for speaker diarization using advanced representation 
    methods (supervectors, speaker factors, i-vectors) and we will describe supervised and 
    unsupervised learning techniques used for speaker diarization. We will also discuss issues 
    such as coping with unknown number of speakers, detecting and dealing with overlapping speech, 
    diarization confidence estimation, and online speaker diarization. Finally we will discuss 
    two recent works: exploiting a-prioiri acoustic information (such as processing a meeting 
    when some of the participants are known in advanced to the system, and training data is available for them), 
    The second recent work is modeling speaker-turn dynamics. If time permits, we will also discuss concepts 
    such as multi-modal diarization and using TDOA (time difference of arrival) for diarization of meetings.

    Organizers: Hagai Aronowitz

T6: Multimodal Speech Recognition with the AusTalk 3D Audio-Visual Corpus

    This tutorial will provide attendees a brief overview of 3D based AVSR research. 
    In this tutorial, attendees will learn how to use the newly developed 3D based audio 
    visual data corpus we derived from the AusTalk corpus ( 
    for audio-visual speech/speaker recognition. In addition, we also plan to introduce 
    some results using this newly developed 3D audio-visual data corpus, which show that 
    there is a significant speech accuracy increase by integrating both depth-level and grey-level 
    visual features. In the first part of the tutorial, we will review some recent works published 
    in the last decade, so that attendees can obtain an overview of the fundamental concepts 
    and challenges in this field. In the second part of the tutorial, we will briefly describe 
    the recording protocol and contents of the 3D data corpus, and show attendees how to use 
    this corpus for their own research. In the third part of this tutorial, we will present our 
    results using the 3D data corpus. The experimental results show that, compared with the 
    conventional AVSR based on the audio and grey-level visual features, the integration of grey 
    and depth visual information can boost the AVSR accuracy significantly. Moreover, 
    we will also experimentally explain why adding depth information can benefit the standard AVSR systems. 
    Eventually, through our tutorial, we hope we can inspire more researchers in the community 
    to contribute to this exciting research.

    Organizers: Roberto Togneri, Mohammed Bennamoun and Chao (Luke) Sui

T7: Semantic Web and Linked Big Data Resources for Spoken Language Processing

    State-of-the-art statistical spoken language processing typically requires 
    significant manual effort to construct domain-specific schemas (ontologies) 
    as well as manual effort to annotate training data against these schemas. 
    At the same time, a recent surge of activity and progress on semantic web-related 
    concepts from the large search-engine companies represents a potential alternative 
    to the manually intensive design of spoken language processing systems. 
    Standards such as have been established for schemas (ontologies) that 
    webmasters can use to semantically and uniformly markup their web pages. 
    Search engines like Bing, Google, and Yandex have adopted these standards and are 
    leveraging them to create semantic search engines at the scale of the web. 
    As a result, the open linked data resources and semantic graphs covering various 
    domains (such as Freebase [3]) have grown massively every year and contains far more 
    information than any single resource anywhere on the Web. Furthermore, these resources 
    contain links to text data (such as Wikipedia pages) related to the knowledge in the graph.

    Recently, several studies on speech language processing started exploiting these massive 
    linked data resources for language modeling and spoken language understanding. 
    This tutorial will include a brief introduction to the semantic web and the linked 
    data structure, available resources, and querying languages. 
    An overview of related work on information extraction and language processing will 
    be presented, where the main focus will be on methods for learning spoken language 
    understanding models from these resources.

    Organizers: Dilek Hakkani-Tür and Larry Heck

T8: Speech and Audio for Multimedia Semantics

    Internet media sharing sites and the one-click upload capability of smartphones 
    are producing a deluge of multimedia content. While visual features are often dominant 
    in such material, acoustic and speech information in particular often complements it. 
    By facilitating access to large amounts of data, the text-based Internet gave a huge 
    boost to the field of natural language processing. The vast amount of consumer-produced 
    video becoming available now will do the same for video processing, eventually enabling 
    semantic understanding of multimedia material, with implications for human computer interaction, robotics, etc.

    Large-scale multi-modal analysis of audio-visual material is now central to a number of 
    multi-site research projects around the world. While each of these have slightly different 
    targets, they are facing largely the same challenges: how to robustly and efficiently process 
    large amounts of data, how to represent and then fuse information across modalities, 
    how to train classifiers and segmenters on unlabeled data, how to include human feedback, etc.

    In this tutorial, we will present the state of the art in large-scale video, speech, 
    and non-speech audio processing, and show how these approaches are being applied to tasks 
    such as content based video retrieval (CBVR) and multimedia event detection (MED). 
    We will introduce the most important tools and techniques, and show how the combination of 
    information across modalities can be used to induce semantics on multimedia material 
    through ranking of information and fusion. Finally, we will discuss opportunities 
    for research that the INTERSPEECH community specifically will find interesting and fertile. 

    Organizers: Florian Metze and Koichi Shinoda

ISCSLP Tutorials @ INTERSPEECH 2014 Description

ISCSLP-T1: Adaptation Techniques for Statistical Speech Recognition

    Adaptation is a technique to make better use of existing models for test data 
    from new acoustic or linguistic conditions. It is an important and challenging 
    research area of statistical speech recognition. This tutorial gives a systematic 
    review of fundamental theories as well as introduction of state-of-the-art adaptation 
    techniques. It includes both acoustic and language model adaptation. Following a simple example 
    of acoustic model adaptation, basic concepts, procedures and categories of adaptation will 
    be introduced. Then, a number of advanced adaptation techniques will be discussed, 
    such as discriminative adaptation, Deep Neural Network adaptation, adaptive training, 
    relationship to noise robustness etc. After the detailed review of acoustic model adaptation, 
    an introduction of language model adaptation, such as topic adaptation will also be given. 
    The whole tutorial is then summarised and future research direction will be discussed.

    Organizers: Kai Yu

ISCSLP-T2: Emotion and Mental State Recognition: Features, Models, System Applications and Beyond

    Emotion recognition is the ability to identify what you are feeling from moment 
    to moment and to understand the connection between your feelings and your expressions. 
    In today’s world, human-computer interaction (HCI) interface undoubtedly plays an 
    important role in our daily life. Toward harmonious HCI interfaces, automated analysis 
    and recognition of human emotion has attracted increasing attention from researchers 
    in multidisciplinary research fields. A specific area of current interest that also has key 
    implications for HCI is the estimation of cognitive load (mental workload), research into 
    which is still at an early stage. Technologies for processing daily activities including speech, 
    text and music have expanded the interaction modalities between humans and computer-supported 
    communicational artifacts.

    In this tutorial, we will present theoretical and practical work offering new and broad views 
    of the latest research in emotional awareness from audio and speech. We discuss several parts 
    spanning a variety of theoretical background and applications ranging from salient emotional features, 
    emotional-cognitive models, compensation methods for variability due to speaker and linguistic content, 
    to machine learning approaches applicable to emotion recognition. In each topic, we will review 
    the state of the art by introducing current methods and presenting several applications. 
    In particular, the application to cognitive load estimation will be discussed, 
    from its psychophysiological origins to system design considerations. Eventually, 
    technologies developed in different areas will be combined for future applications, 
    so in addition to a survey of future research challenges, 
    we will envision a few scenarios in which affective computing can make a difference.

    Organizers: Chung-Hsien Wu, Hsin-Min Wang, Julien Epps and Vidhyasaharan Sethu

ISCSLP-T3: Unsupervised Speech and Language Processing via Topic Models

    In this tutorial, we will present state-of-art machine learning approaches 
    for speech and language processing with highlight on the unsupervised methods 
    for structural learning from the unlabeled sequential patterns. In general, 
    speech and language processing involves extensive knowledge of statistical models. 
    We require designing a flexible, scalable and robust system to meet heterogeneous 
    and nonstationary environments in the era of big data. This tutorial starts from an 
    introduction of unsupervised speech and language processing based on factor analysis 
    and independent component analysis. The unsupervised learning is generalized to a latent 
    variable model which is known as the topic model. The evolution of topic models from 
    latent semantic analysis to hierarchical Dirichlet process, from non-Bayesian parametric 
    models to Bayesian nonparametric models, and from single-layer model to hierarchical 
    tree model shall be surveyed in an organized fashion. The inference approaches based on 
    variational Bayesian and Gibbs sampling are introduced. We will also present several 
    case studies on topic modeling for speech and language applications including language model, 
    document model, retrieval model, segmentation model and summarization model. 
    At last, we will point out new trends of topic models for speech and language processing.

    Organizers: Jen-Tzung Chien

ISCSLP-T4: Deep Learning for Speech Generation and Synthesis

    Deep learning, which can represent high-level abstractions in data with an architecture of 
    multiple non-linear transformation, has made a huge impact on automatic speech recognition (ASR) 
    research, products and services. However, deep learning for speech generation and synthesis 
    (i.e., text-to-speech), which is an inverse process of speech recognition (i.e., speech-to-text), 
    has not generated the similar momentum as it is for ASR yet. Recently, motivated by the success 
    of Deep Neural Networks in speech recognition, some neural network based research attempts have 
    been tried successfully on improving the performance of statistical parametric based 
    speech generation/synthesis. In this tutorial, we focus on deep learning approaches to the 
    problems in speech generation and synthesis, especially on Text-to-Speech (TTS) synthesis and voice conversion.

    First, we give a review for the current main stream of statistical parametric based speech generation 
    and synthesis, or the GMM-HMM based speech synthesis and GMM-based voice conversion with emphasis 
    on analyzing the major factors responsible for the quality problems in the GMM-based voice 
    synthesis/conversion and the intrinsic limitations of a decision-tree based, contextual state 
    clustering and state-based statistical distribution modeling. We then present the latest deep 
    learning algorithms for feature parameter trajectory generation, in contrast to deep learning for 
    recognition or classification. We cover common technologies in Deep Neural Network (DNN) and improved 
    DNN: Mixture Density Networks (MDN), Recurrent Neural Networks (RNN) with Bidirectional Long Short 
    Term Memory (BLSTM) and Conditional RBM (CRBM). Finally, we share our research insights and hand-on 
    experience on building speech generation and synthesis systems based upon deep learning algorithms.

    Organizers: Yao Qian and Frank K. Soong

3-1-11(2014-09-14) INTERSPEECH 2014 Grants and awards

--- September 14-18, 2014

INTERSPEECH 2014 Organizing Committee is glad to announce that
5 travel grants will be offered by INTESPEECH 2014 sponsors and
that ISCA will offer 55 grants for students and young scientists
and give 3 ISCA Best Student Paper Awards.
Modalities and details about the grants and Best Paper Awards
are given below.

-- ISCA Grants

General Information

    ISCA will offer grants to students and young scientists (age<35)
    in order to help them attend the conference. Each grant, equivalent
    to 650 Euros, will be disbursed in cash at the registration desk
    during the conference.

    All grant recipients are strongly encouraged to participate
    in the Student Volunteer Program with a workload of one day
    or 3 sessions for the entire conference. For details - please
    refer to the Student Volunteer Program page.


    The grants for participation in INTERSPEECH 2014 will be
    administered by the ISCA Board and awarded, according to the
    available budget, to applicants who provide the information
    listed below:

        * Letter of acceptance of a paper to be presented.
        * Mentioning of other funding being requested,
          awarded or rejected by other sources
        * Sworn statement that the applicant never received any
          previous grants from ISCA.
    Note that preference is given to applicants from areas
    needing greater support.

    The purpose of these grants is to guarantee a representation
    of the diverse scientific topics of the conference and a cross-
    section of the international speech community as wide as possible.
    Exceptionally, researchers in special situations like unemployment
    or coming from low income-level countries could also apply.
    Applicants from all countries are eligible.

Application Process

    Applicants should complete the application form with
    additional documents:

        1) Curriculum Vitae with academic records
        2) Complete list of previous publications
           (title, year, co-authors journal/book/proceedings, pages).
        3) List of the conferences and workshops previously attended.
        4) Letter of acceptance of the paper(s) from the conference organizers
        5) Copy of the original manuscript(s) accepted to the conference
        6) Recommendation letter (email is acceptable)
           from the Director of your Laboratory

    Application form can be downloaded from

    Please submit the application package by sending an email with
    subject heading “ISCA Grants for INTERSPEECH 2014” to the grants
    coordinator Prof. Alan Black:

    before 17 June 2014. Prof. Alan Black will send notification letter
    to all applicants.For more information of grant application,
    please refer to

    As there is a limited period for application,
    interested applicants are encouraged to start preparing the
    required documents if s/he would like to consider this source
    of support to attend INTERSPEECH 2014.

*                         Important Dates                              *
Grant application start after acceptance notification:     June 10, 2014
Grant application deadline:                                June 17, 2014
Notification of grant acceptance:                          June 24, 2014
Early Registration:                                        July 10, 2014

Registration Procedure

    All grant recipients must be ISCA members to receive the grants.
    All grant recipients will be given free registration to the
    INTERSPEECH main conference that includes the technical program,
    reception, and banquet, but excludes other options (such as
    tutorials). Grant recipients please inform the organizer your
    Registration ID, which you can obtain from the online registration
    system, before you make payment (if any) so that the organizer can
    waive your registration fee by changing your account setting.

    All grants will be disbursed onsite, during the conference.
    Please see Dr Tong Rong with the notification letter and paper ID
    at the registration desk, to collect your money.

    All grant recipients are required to register online to the system
    to confirm the attendance, to make purchases beyond the free
    package. As all accounts have to be settled immediately, the
    organizers will not entertain any request to transfer, ship or send
    money after the conference ends.

    ISCA Board will make the selection and provide the list of grant
    recipients to INTERSPEECH 2014 organizing committee. The committee
    will inform you the registration procedure. Please contact,

    if you have any inquiry. Please note that grant recipients need to
    complete the registration by 10 Jul 2014. Late registration fees
    apply if this early registration deadline is not met.

-- INTERSPEECH 2014 Student Travel Grants

    INTERSPEECH 2014 would like to express sincere gratitude to Google,
    Samsung and IFLYTEK CO., LTD. for their combined support of
    5 student travel grants, 3 will be selected as the Google grantees,
    1 will be selected as the Samsung grantee, and 1 will be selected
    as the IFLYTEK CO., LTD. grantee. The grantees are selected based
    on the technical quality of their paper(s). The grantees will be
    notified and announced on the INTERSPEECH 2014 website by July 3,
    2014. Grantees will be announced by the INTERSPEECH General Chair
    during the closing ceremony.

    For any inquiry, please contact

ISCA Best Student Paper Awards

    Each year, ISCA awards 3 best student paper awards at INTERSPEECH
    based on the technical merit of the paper and the presentation at
    the conference. Note that the first author student needs to present
    the paper in person to be considered for the award. The ISCA Best
    Student Paper Award is sponsored by IBM Research. In addition,
    Dilek Hakkani-Tur and Larry Heck from Microsoft Research are
    donating their honorarium from their tutorial “Semantic Web and
    Linked Big Data Resources for Spoken Language Processing” to
    financially sponsor the ISCA Best Student Paper Award. The
    shortlist of Best Student Paper Awards will be announced on the
    INTERSPEECH 2014 website by July 3, 2014 and the final awardees
    will be announced by the president of ISCA during the closing

    For any inquiry, please contact

The INTERSPEECH 2014 Organizing Committee


3-1-12(2014-09-14) INTERSPEECH 2014 News (June 2014)

INTERSPEECH 2014 satellite workshops will take place in Singapore, Phuket (Thailand) and Penang (Malaysia) for more speech events before or after #is2014



Join INTERSPEECH 2014 satellite 2nd workshop on Multimodal Analyses Enabling Artificial Agents in Human-Machine Interaction: MA3HMI 2014


Enter the Blizzard Challenge Workshop 2014 in Singapore right after INTERSPEECH 2014. More info on


After INTERSPEECH 2014, be welcome to join the4th Workshop on Child, Computer and Interaction that will be hel in Singapore on September 19th , 2014. More info at


Do you know that the 17th Oriental COCOSDA will take place in Phuket, Thailand before INTERSPEECH 2014? more at


The 2nd Workshop on Speech, Language and Audio in Multimedia (SLAM) is organized in Penang Malaysia as INTERSPEECH 2014 satellite event.


Register for INTERSPEECH 2014 tutorial on “Non-speech acoustic event detection and classification”

Learn about the “Contribution of MRI to Exploring and Modeling Speech Production” during INTERSPEECH 2014 tutorials


Find out about what's new in “Computational Models for Audiovisual Emotion Perception” in INTERSPEECH 2014 tutorials.


Join the tutorial on “The Art and Science of Speech Feature Engineering” before INTERSPEECH 2014.


What are the “Recent Advances in Speaker Diarization”? Come and learn about them in INTERSPEECH 2014 tutorials.


“Multimodal Speech Recognition with the AusTalk 3D Audio-Visual Corpus” will be the subject of a tutorial of INTERSPEECH 2014.


Please check out INTERSPEECH 2014 tutorial on “Semantic Web and Linked Big Data Resources for Spoken Language Processing”.


Participate to the tutorial on “Speech and Audio for Multimedia Semantics” during INTERSPEECH 2014.


Join ISCSLP@INTERSPEECH tutorial on 'Adaptation Techniques for Statistical Speech Recognition” in INTERSPEECH 2014 satellite event


Register to ISCSLP@INTERSPEECH tutorial on “Emotion and Mental State Recognition: Features, Models, System Applications and Beyond”,


Learn more about “Unsupervised Speech and Language Processing via Topic Models”,in ISCSLP@INTERSPEECH tutorial


ISCSLP@ INTERSPEECH will host a tutorial on “Deep Learning for Speech Generation and Synthesis”,




3-1-13(2014-09-14) INTERSPEECH 2014 Plenary talks

-            INTERSPEECH 2014 - SINGAPORE           -
-               September 14-18, 2014               -
-           -

ISCA, COLIPS and the organizing Committee of INTERSPEECH 2014
are proud to announce that INTERSPEECH 2014 will feature
five plenary talks by internationally renowned experts.

- keynote speech
  by the ISCA Medallist 2014

- 'Decision Learning in Data Science:
  Where John Nash Meets Social Media'
  by Professor K. J. Ray Liu

- 'Language Diversity: Speech Processing In A Multi-Lingual Context'
  by Dr. Lori Lamel

- 'Sound Patterns In Language'
  by Professor William Shi-Yuan WANG 王士元

- 'Achievements and Challenges of Deep Learning
  From Speech Analysis And Recognition To Language
  And Multimodal Processing'
  by Dr. Li DENG

Details of the keynote speeches and biographies of the presenters are given below.

Looking forward to welcome you in Singapore,
the organizing committee

* On Monday, 15th of September                                         *

The ISCA Medallist 2014 will give a keynote speech.
The name of the Medallist and subject of the talk will be
disclosed on the first day of INTERSPEECH 2014.

* On Tuesday morning, 16th of September                                *

Professor K. J. Ray Liu
Department of Electrical and Computer Engineering
University of Maryland, College Park

will give a presentation on:

'Decision Learning in Data Science: Where John Nash Meets Social Media'


    With the increasing ubiquity and power of mobile devices,
    as well as the prevalence of social media, more and more
    activities in our daily life are being recorded, tracked,
    and shared, creating the notion of “social media”.
    Such abundant and still growing real life data, known as
    “big data”, provide a tremendous research opportunity in many fields.
    To analyze, learn and understand such user-generated big data,
    machine learning has been an important tool and various
    machine learning algorithms have been developed.
    However, since the user-generated big data is
    the outcome of users’ decisions, actions and their socio-economic
    interactions, which are highly dynamic, without considering users’
    local behaviours and interests, existing learning approaches
    tend to focus on optimizing a global objective function at
    the macroeconomic level, while totally ignore users’ local
    decisions at the micro-economic level. As such there is a growing
    need in bridging machine/social learning with strategic decision
    making, which are two traditionally distinct research disciplines,
    to be able to jointly consider both global phenomenon and local
    effects to understand/model/analyze better the newly arising
    issues in the emerging social media. In this talk, we present
    the notion of “decision learning” that can involve users's
    behaviours and interactions by combining learning with strategic
    decision making.
    We will discuss some examples from social media with real data to
    show how decision learning can be used to better analyze users’
    optimal decision from a user’ perspective as well as design a
    mechanism from the system designer’s perspective
    to achieve a desirable outcome.

Biography of the speaker

    Dr. K. J. Ray Liu was named a Distinguished Scholar-Teacher
    of University of Maryland in 2007, where he is Christine Kim
    Eminent Professor of Information Technology.
    He leads the Maryland Signals and Information Group conducting
    research encompassing broad areas of signal processing and
    communications with recent focus on cooperative communications,
    cognitive networking, social learning and decision making,
    and information forensics and security. Dr. Liu has received
    numerous honours and awards including IEEE Signal Processing
    Society 2009 Technical Achievement Award and various best paper
    awards from IEEE Signal Processing, Communications, and Vehicular
    Technology Societies, and EURASIP. A Fellow of the IEEE and AAAS,
    he is recognized by Thomson Reuters as an ISI Highly Cited
    Dr. Liu was the President of IEEE Signal Processing Society,
    the Editor-in-Chief of IEEE Signal Processing Magazine and
    the founding Editor-in-Chief of EURASIP Journal on Advances
    in Signal Processing. Dr. Liu also received various research
    and teaching recognitions from the University of Maryland,
    including Poole and Kent Senior Faculty Teaching Award,
    Outstanding Faculty Research Award, and Outstanding Faculty
    Service Award, all from A. James Clark School of Engineering;
    and Invention of the Year Award (three times)
    from Office of Technology Commercialization.

* On Tuesday afternoon, 16th of September                              *

Dr. Lori Lamel
Senior Research scientist (DR1), LIMSI-CNRS

will give a presentation on

'Language Diversity: Speech Processing In A Multi-Lingual Context'


    Speech processing encompasses a variety of technologies
    that automatically process speech for some downstream processing.
    These technologies include identifying the language or dialect
    spoken, the person speaking, what is said and how it is said.
    The downstream processing may be limited to a transcription or
    to a transcription enhanced with additional meta-data, or may
    be used to carry out an action or interpreted within a spoken
    dialogue system or more generally for analytics.  With the
    availability of large spoken multimedia or multimodal data there is
    growing interest in using such technologies to provide structure
    and random access to particular segments. Automatic tools can also
    serve to annotate large corpora for exploitation in linguistic
    studies of spoken language, such as acoustic-phonetics,
    pronunciation variation and diachronic evolution,
    permitting the validation of hypotheses and models.
    In this talk I will present some of my experience with speech
    processing in multiple languages, drawing upon progress in the
    context of several research projects, most recently the Quaero
    program and the IARPA Babel program, both of which address the
    development of technologies in a variety of languages, with the aim
    to some highlight recent research directions and challenges.

Biography of the speaker

    I am a senior research scientist (DR1) at the CNRS, which I joined as
    a permanent researcher at LIMSI in October 1991.
    I received my Ph.D. degree in Electrical Engineering and Computer Science
    in May 1988 from the Massachusetts Institute of Technology.
    My research activities focus on large vocabulary speaker-
    independent, continuous speech recognition in multiple languages
    with a recent focus on low-resourced languages; lightly and
    unsupervised acoustic model training methods; studies in acoustic-
    phonetics; lexical and pronunciation modelling. I contributed to
    the design, and realization of large speech corpora (TIMIT, BREF,
    TED). I have been actively involved in the research projects, most
    recently leading the activities on speech processing in the OSEO
    Quaero program, and I am currently co-principal investigator for
    LIMSI as part of the IARPA Babel Babelon team led by BBN.
    I served on the Steering committee for Interspeech 2013 as
    co-technical program chair along with Pascal Perrier, and I am now
    serving on the Technical Program Committee of Interspeech 2014.

* On Wednesday, 17th of September                                      *

Professor William Shi-Yuan WANG 王士元
Centre for Language and Human Complexity,
Chinese University of Hong Kong
Professor Emeritus, University of California at Berkeley
Honorary Professor, Peking University
Academician, Academia Sinica

will give a presentation about

'Sound Patterns In Language'


    In contrast to other species, humans are unique in having developed
    thousands of diverse languages which are not mutually
    intelligible. However, any infant can learn any language with ease,
    because all languages are based upon common biological
    infrastructures of sensori-motor, memorial, and cognitive
    faculties.  While languages may differ significantly in the sounds
    they use, the overall organization is largely the same.
    It is divided into a discrete segmental system for building words
    and a continuous prosodic system for expressing, phrasing,
    attitudes, and emotions. Within this organization, I will discuss a
    class of languages called 'tone languages', which makes special use
    of F0 to build words.  Although the best known of these is Chinese,
    tone languages are found in many parts of the world, and operate on
    different principles. I will also comment on relations between
    sound patterns in language and sound patterns in music, the two
    worlds of sound universal to our species.

Biography of the speaker

    William S-Y. Wang received his early schooling in China, and his
    PhD from the University of Michigan.  He was appointed
    Professor of Linguistics at the University of California at
    Berkeley in 1965, and taught there for 30 years.
    Currently he is in the Department of Electronic Engineering and in
    the Department of Linguistics and Modern Languages of the Chinese
    University of Hong Kong, and Director of the newly established
    Joint Research Centre for Language and Human Complexity. His
    primary interest is the evolution of language from a multi-
    disciplinary perspective.

* On Thursday, 18th of September                                      *

Principal Researcher and Research Manager
Deep Learning Technology Centre,
Microsoft Research, Redmond, USA

will give a presentation on the

'Achievements and Challenges of Deep Learning
From Speech Analysis And Recognition To Language And Multimodal Processing'


    Artificial neural networks have been around for over half a century
    and their applications to speech processing have been almost as
    long, yet it was not until year 2010 that their real impact had
    been made by a deep form of such networks, built upon part of the
    earlier work on (shallow) neural nets and (deep) graphical models
    developed by both speech and machine learning communities. This
    keynote will first reflect on the path to this transformative
    success, sparked by speech analysis using deep learning methods
    on spectrogram-like raw features and then progressing rapidly to
    speech recognition with increasingly larger vocabularies and scale.
    The role of well-timed academic-industrial collaboration will be
    highlighted, so will be the advances of big data, big compute, and
    the seamless integration between the application-domain knowledge
    of speech and general principles of deep learning. Then, an
    overview will be given on sweeping achievements of deep learning in
    speech recognition since its initial success in 2010 (as well as in
    image recognition and computer vision since 2012). Such
    achievements have resulted in across-the-board, industry-wide
    deployment of deep learning. The final part of the talk will look
    ahead towards stimulating new challenges of deep learning ---
    making intelligent machines capable of not only hearing (speech)
    and seeing (vision), but also of thinking with a “mind”; i.e.
    reasoning and inference over complex, hierarchical relationships
    and knowledge sources that comprise a vast number of entities
    and semantic concepts in the real world based in part on multi-
    sensory data from the user.  To this end, language and multimodal
    processing --- joint exploitation and learning from text,
    speech/audio, and image/video --- is evolving into a new frontier
    of deep learning, beginning to be embraced by a mixture of research
    communities including speech and spoken language processing,
    natural language processing, computer vision, machine learning,
    information retrieval, cognitive science, artificial intelligence,
    and data/knowledge management. A review of recent published studies
    will be provided on deep learning applied to selected language and
    multimodal processing tasks, with a trace back to the relevant
    early connectionist modelling and neural network literature and
    with future directions in this new exciting deep learning frontier
    discussed and analyzed.

Biography of the speaker

    Li Deng received Ph.D. from the University of Wisconsin-Madison.
    He was a tenured professor (1989-1999) at the University of
    Waterloo, Ontario, Canada, and then joined Microsoft Research,
    Redmond, where he is currently a Principal Research Manager of its
    Deep Learning Technology Centre.
    Since 2000, he has also been an affiliate full professor at the
    University of Washington, Seattle, teaching computer speech
    processing. He has been granted over 60 US or international
    patents, and has received numerous awards and honours
    bestowed by IEEE, ISCA, ASA, and Microsoft including the latest
    IEEE SPS Best Paper Award (2013) on deep neural nets for speech
    recognition. He authored or co-authored 4 books including the
    latest one on Deep Learning: Methods and Applications. He is a
    Fellow of the Acoustical Society of America, a Fellow of the IEEE,
    and a Fellow of the ISCA. He served as the Editor-in-Chief
    for IEEE Signal Processing Magazine (2009-2011), and currently as
    Editor-in-Chief for IEEE Transactions on Audio, Speech and Language
    Processing. His recent research interests and activities have been
    focused on deep learning and machine intelligence applied to
    large-scale text analysis and to speech/language/image
    multimodal processing, advancing his earlier work with
    collaborators on speech analysis and recognition using deep neural
    networks since 2009.


3-1-14(2014-09-14) INTERSPEECH 2014 Singapore



It is a great pleasure to announce that the 15th edition of the Annual Conference of the International Speech Communication Association (INTERSPEECH) will be held in Singapore during September 14-18, 2014. INTERSPEECH 2014 will bring together the community to celebrate the diversity of spoken languages in the vibrant city state of Singapore.  INTERSPEECH 2014 is proudly organized by the Chinese and Oriental Languages Information Processing Society (COLIPS), the Institute for Infocomm Research (I2R), and the International Speech Communication Association (ISCA).




Ten steps to Singapore


You want to know more about Singapore?


During the next ten months, the organization committee will introduce you to Singaporean culture through a series of brief newsletters featuring topics related to spoken languages in Singapore. Please stay tuned!






Submission deadline:  December 1, 2013


Satellite workshops related to speech and language research will be hosted in Singapore as well as in Phuket Island, Thailand (1 hr 20 min flight from Singapore) and in Penang, Malaysia (1 hr flight from Singapore).


Proposals must be submitted by email to before December 1, 2013. Notification of acceptance and ISCA approval/sponsorship will be announced by January 31, 2014.




Sponsorship and Exhibition


The objective of INTERSPEECH 2014 is to foster scientific exchanges in all aspects of Speech Communication sciences with a special focus on the diversity of spoken languages. We are pleased to invite you to take part in this major event as a sponsor. For more information, view the Sponsorship




Conference venue


INTERSPEECH 2014 main conference will be held in the MAX Atria @ Singapore Expo.






Lists of the organizing, advisory and technical program committees are available on line (here).




Follow us


Facebook: ISCA


Twitter: @Interspeech2014 follow hash tags: #is2014 or #interspeech2014


LinkedIn Interspeech






Conference website:— For general enquiries — For Exhibition & Sponsorship — For Workshops & Satellite Events








3-1-15(2014-09-14) Interspeech 2014 special session : Speech technologies for Ambient Assisted Living.

Interspeech 2014 special session : Speech technologies for Ambient Assisted Living.

Submission deadline: 24th March 2014

Singapore, 14-18 September 2014 This special session focuses on the use of speech technologies for ambient assisted living, the creation of smart spaces and intelligent companions that can preserve independence and executive function, social communication and security of people with special needs. Currently, speech input for assistive technologies remains underutilized despite its potential to deliver highly informative data and serve as the primary means of interaction with the home. Speech interfaces could replace or augment obtrusive and sometimes outright inaccessible conventional computer interfaces. Moreover, the smart home context can support speech communication by providing a number of concurrent information sources (e.g., wearable sensors, home automation sensors, etc.), enabling multimodal communication. In practice, its use remains limited due to challenging real-world conditions, and because conventional speech interfaces can have difficulty with the atypical speech of many users. This, in turn, can be attributed to the lack of abundant speech material, and the limited adaptation to the user of these systems. Taking up the challenges of this domain requires a multidisciplinary approach to define the user's needs, record corpora in realistic usage conditions, develop speech interfaces that are robust to both environment and user's characteristics and are able to adapt to specific users. This special session aims at bringing together researchers in speech and audio technologies with people from the ambient assisted living and assistive technologies communities to meet and foster awareness between members of either community, discuss problems, techniques and datasets, and perhaps initiate common projects. Topics of the session will include: Assistive speech technology Applications of speech technology (ASR, dialogue, synthesis) for ambient assisted living Understanding, modelling, or recognition of aged and atypical speech Multimodal speech recognition (context-aware ASR) Multimodal emotion recognition Audio scene and smart space context analysis Assessment of speech and language processing within the context of assistive technology Speech synthesis and speech recognition for physical or cognitive impairments Symbol languages, sign languages, nonverbal communication Speech and NLP applied to typing interface applications Language modelling for Augmentative and Alternative Communication text entry and speech generating devices Deployment of speech and NLP tools in the clinic or in the field Linguistic resources; corpora and annotation schemes Evaluation of systems and components. Submission instructions: Researchers who are interested in contributing to this special session are invited to submit a paper according to the regular submission procedure of INTERSPEECH 2014, and to select 'Speech technologies for Ambient Assisted Living' in the special session field of the paper submission form. Please feel free to contact the organisers if you have any question regarding the special session.

Organizers: Michel Vacher michel.vacher [at] Laboratoire d'Informatique de Grenoble, François Portet francois.portet [at] Laboratoire d'Informatique de Grenoble, Frank Rudzicz frank [at] University of Toronto, Jort F. Gemmeke jgemmeke [at] KU Leuven, Heidi Christensen h.christensen [at] University of Sheffield,


3-1-16(2014-09-14) Invitation to submit at the INTERSPEECH 2014 workshops (still open for only one of them)


To  INTERSPEECH 2014 Authors:

You should have received the INTERSPEECH 2014 notification of paper acceptance by now.

We are glad to let you know that 6 INTERSPEECH Workshops/Satellite Workshops are co-located with INTERSPEECH 2014. Their paper submissions are still open. You are encouraged to submit papers to the workshops and to join the workshops as part of your INTERSPEECH trip.

The details of the workshops are as follows.

ISCSLP 2014: 9th International Symposium on Chinese Spoken Language Processing
Date: 12-14 September 2014
Location: Singapore
Submission Deadline: 17 June 2014 (special session - Advances in Human Language Technologies)

WOCCI 2014: 4th Workshop on Child, Computer and Interaction
Date: 19 September 2014
Location: Singapore
Submission Deadline: 17 June 2014

O-COCOSDA 2014: 17th Oriental COCOSDA
Date: 10-12 September 2014
Location: Phuket, Thailand
Submission Deadline: 20 June 2014

SLAM 2014: 2nd Workshop on Speech, Language and Audio in Multimedia
Date: 11-12 September 2014
Location: Penang, Malaysia
Submission Deadline: 20 June 2014

MA3HMI 2014: 2nd Workshop on Multimodal Analyses Enabling Artificial Agents in Human-Machine Interaction
Date: 14 September 2014
Location: Singapore
Submission Deadline: 1 July 2014

Blizzard Challenge Workshop 2014
Date: 19 September 2014
Location: Singapore
Submission Deadline: 9 August 2014

Yours sincerely,
Helen Meng and Bin Ma
Technical Program Chairs, INTERSPEECH 2014

3-1-17(2014-09-14) Special sessions at Interspeech 2014: call for submissions


--- September 14-18, 2014


INTERSPEECH is the world's largest and most comprehensive conference on issues surrounding

the science and technology of spoken language processing, both in humans and in machines.

The theme of INTERSPEECH 2014 is

--- Celebrating the Diversity of Spoken Languages ---

INTERSPEECH 2014 includes a number of special sessions covering interdisciplinary topics

and/or important new emerging areas of interest related to the main conference topics.

Special sessions proposed for the forthcoming edition are:

• A Re-evaluation of Robustness

• Deep Neural Networks for Speech Generation and Synthesis

• Exploring the Rich Information of Speech Across Multiple Languages

• INTERSPEECH 2014 Computational Paralinguistics ChallengE (ComParE)

• Multichannel Processing for Distant Speech Recognition

• Open Domain Situated Conversational Interaction

• Phase Importance in Speech Processing Applications

• Speaker Comparison for Forensic and Investigative Applications

• Text-dependent for Short-duration Speaker Verification

• Tutorial Dialogues and Spoken Dialogue Systems

• Visual Speech Decoding

A description of each special session is given below.

For paper submission, please follow the main conference procedure and chose the Special Session track when selecting

your paper area.

Paper submission procedure is described at:

For more information, feel free to contact the Special Session Chair,

Dr. Tomi H. Kinnunen, at email tkinnu [at]


Special Session Description


A Re-evaluation of Robustness

The goal of the session is to facilitate a re-evaluation of robust speech

recognition in the light of recent developments. It’s a re-evaluation at two levels:

• a re-evaluation in perspective brought by breakthroughs in performance obtained

by Deep Neural Network which leads to a fresh questioning of the role and

contribution of robust feature extraction.

• A literal re-evaluation on common databases to be able to present and compare

performances of different algorithms and system approaches to robustness.

Paper submissions are invited on the theme of noise robust speech recognition

and required to submit results on the Aurora 4 database to facilitate cross comparison

of the performance between different techniques.

Recent developments raise interesting research questions that the session aims to help

Progress by bringing focus and exploration of these issues. For example

1. What role is there for signal processing to create feature representations to use as

inputs to Deep Learning or can deep learning do all the work?

2. What feature representations can be automatically learnt in a deep learning architecture?

3. What other techniques can give great improvement in robustness?

4. What techniques don’t work and why?

The session organizers wish to encourage submissions that bring insight and understanding to

the issues highlighted above. Authors are requested not only to present absolute performance

of the whole system but also to highlight the contribution made by various components in a

complex system.

Papers that are accepted for the session are encouraged to also evaluate their techniques on new test

data sets (available in July) and submit their results at the end of August.

Session organization

The session will be structured as a combination of

1. Invited talks

2. Oral paper presentations

3. Poster presentations

4. Summary of contributions and results on newly released test sets

5. Discussion


David Pearce, Audience dpearce [at]

Hans-Guenter Hirsch, Niederrhein University of Applied Sciences, hans-guenter.hirsch [at]

Reinhold Haeb-Umbach, University of Paderborn, haeb [at]

Michael Seltzer, Microsoft, mseltzer [at]

Keikichi Hirose, The University of Tokyo, hirose [at]

Steve Renals, University of Edinburgh, s.renals [at]

Sim Khe Chai, National University of Singapore, simkc [at]

Niko Moritz, Fraunhofer IDMT, Oldenburg, niko.moritz [at]

K K Chin, Google, kkchin [at]

Deep Neural Networks for Speech Generation and Synthesis

This special session aims to bring together researchers who work actively on deep neural

networks for speech research, particularly, in generation and synthesis, to promote and

to understand better the state-of-art DNN research in statistical learning and compare

results with the parametric HMM-GMM model which has been well-established for speech synthesis,

generation, and conversion. DNN, with its neuron-like structure, can simulate human speech

production system in a layered, hierarchical, nonlinear and self-organized network.

It can transform linguistic text information into intermediate semantic, phonetic and prosodic

content and finally generate speech waveforms. Many possible neural network architectures or

typologies exist, e.g. feed-forward NN with multiple hidden layers, stacked RBM or CRBM,

Recurrent Neural Net (RNN), which have been used to speech/image recognition and other applications.

We would like to use this special session as a forum to present updated results in the research frontiers,

algorithm development and application scenarios. Particular focused areas will be on

parametric TTS synthesis, voice conversion, speech compression, de-noising and speech enhancement.


Yao Qian, Microsoft Research Asia, yaoqian [at]

Frank K. Soong, Microsoft Research Asia, frankkps [at]

Exploring the Rich Information of Speech Across Multiple Languages

Spoken language is the most direct means of communication between human beings. However,

speech communication often demonstrates its language-specific characteristics because of,

for instance, the linguistic difference (e.g., tonal vs. non-tonal, monosyllabic vs. multisyllabic)

across languages. Our knowledge on the diversities of speech science across languages is still limited,

including speech perception, linguistic and non-linguistic (e.g., emotion) information, etc.

This knowledge is of great significance to facilitate our design of language-specific application of

speech techniques (e.g., automatic speech recognition, assistive hearing devices) in the future.

This special session will provide an opportunity for researchers from various communities

(including speech science, medicine, linguistics and signal processing) to stimulate further discussion

and new research in the broad cross-language area, and present their latest research on understanding

the language-specific features of speech science and their applications in the speech communication of

machines and human beings. This special session encourages contributions all fields on speech science,

e.g., production and perception, but with a focus on presenting the language-specific characteristics

and discussing their implications to improve our knowledge on the diversities of speech science across

multiple languages. Topics of interest include, but are not limited to:

1. characteristics of acoustic, linguistic and language information in speech communication across

multiple languages;

2. diversity of linguistic and non-linguistic (e.g., emotion) information among multiple spoken languages;

3. language-specific speech intelligibility enhancement and automatic speech recognition techniques; and

4. comparative cross-language assessment of speech perception in challenging environments.


Junfeng Li, Institute of Acoustics, Chinese Academy of Sciences, [at]

Fei Chen, The University of Hong Kong, feichen1 [at]

INTERSPEECH 2014 Computational Paralinguistics ChallengE (ComParE)

The INTERSPEECH 2014 Computational Paralinguistics ChallengE (ComParE) is an open Challenge

dealing with speaker characteristics as manifested in their speech signal's acoustic properties.

This year, it introduces new tasks by the Cognitive Load Sub-Challenge, the Physical Load

Sub-Challenge, and a Multitask Sub-Challenge: For these Challenge tasks,


and the ANXIETY-DEPRESSION-EMOTION-SLEEPINESS audio corpus (ADES) with high diversity of

speakers and different languages covered (Australian English and German) are provided by the organizers.

All corpora provide fully realistic data in challenging acoustic conditions and feature rich

annotation such as speaker meta-data. They are given with distinct definitions of test,

development, and training partitions, incorporating speaker independence as needed in most

real-life settings. Benchmark results of the most popular approaches are provided as in the years before.

Transcription of the train and development sets will be known. All Sub-Challenges allow contributors

to find their own features with their own machine learning algorithm. However, a standard feature set

will be provided per corpus that may be used. Participants will have to stick to the definition of

training, development, and test sets. They may report on results obtained on the development set,

but have only five trials to upload their results on the test sets, whose labels are unknown to them.

Each participation will be accompanied by a paper presenting the results that undergoes peer-review

and has to be accepted for the conference in order to participate in the Challenge.

The results of the Challenge will be presented in a Special Session at INTERSPEECH 2014 in Singapore.

Further, contributions using the Challenge data or related to the Challenge but not competing within

the Challenge are also welcome.

More information is given also on the Challenge homepage:


Björn Schuller, Imperial College London / Technische Universität München,schuller [at]

Stefan Steidl, Friedrich-Alexander-University, stefan.steidl [at]

Anton Batliner, Technische Universität München / Friedrich-Alexander-University,

batliner [at]

Jarek Krajweski, Bergische Universität Wuppertal, krajewsk [at]

Julien Epps, The University of New South Wales / National ICT Australia, j.epps [at]

Multichannel Processing for Distant Speech Recognition

Distant speech recognition in real-world environments is still a challenging problem: reverberation

and dynamic background noise represent major sources of acoustic mismatch that heavily decrease ASR

performance, which, on the contrary, can be very good in close-talking microphone setups.

In this context, a particularly interesting topic is the adoption of distributed microphones for

the development of voice-enabled automated home environments based on distant-speech interaction:

microphones are installed in different rooms and the resulting multichannel audio recordings capture

multiple audio events, including voice commands or spontaneous speech, generated in various locations

and characterized by a variable amount of reverberation as well as possible background noise.

The focus of the proposed special session will be on multichannel processing for automatic speech recognition (ASR)

in such a setting. Unlike other robust ASR tasks, where static adaptation or training with noisy data sensibly

ameliorates performance, the distributed microphone scenario requires full exploitation of multichannel

information to reduce the highly variable dynamic mismatch. To facilitate better evaluation of the proposed

algorithms the organizers will provide a set of multichannel recordings in a domestic environment.

The recordings will include spoken commands mixed with other acoustic events occurring in different

rooms of a real apartment.

The data is being created in the frame of the EC project DIRHA (Distant speech Interaction for Robust

Home Applications)

which addresses the challenges of speech interaction for home automation.

The organizers will release the evaluation package (datasets and scripts) on February 17;

the participants are asked to submit a regular paper reporting speech recognition results

on the evaluation set and comparing their performance with the provided reference baseline.

Further details are available at:


Marco Matassoni, Fondazione Bruno Kessler, matasso [at]

Ramon Fernandez Astudillo, Instituto de Engenharia de Sistemas e Computadores, ramon.astudillo [at]

Athanasios Katsamanis, National Technical University of Athens, nkatsam [at]

Open Domain Situated Conversational Interaction

Robust conversational systems have the potential to revolutionize our interactions with computers.

Building on decades of academic and industrial research, we now talk to our computers, phones,

and entertainment systems on a daily basis. However, current technology typically limits conversational

interactions to a few narrow domains/topics (e.g., weather, traffic, restaurants). Users increasingly want

the ability to converse with their devices over broad web-scale content. Finding something on your PC or

the web should be as simple as having a conversation.

A promising approach to address this problem is situated conversational interaction. The approach leverages

the situation and/or context of the conversation to improve system accuracy and effectiveness.

Sources of context include visual content being displayed to the user, Geo-location, prior interactions,

multi-modal interactions (e.g., gesture, eye gaze), and the conversation itself. For example, while a user

is reading a news article on their tablet PC, they initiate a conversation to dig deeper on a particular topic.

Or a user is reading a map and wants to learn more about the history of events at mile marker 121.

Or a gamer wants to interact with a game’s characters to find the next clue in their quest.

All of these interactions are situated – rich context is available to the system as a source of priors/constraints

on what the user is likely to say.

This special session will provide a forum to discuss research progress in open domain situated

conversational interactions.

Topics of the session will include:

• Situated context in spoken dialog systems

• Visual/dialog/personal/geo situated context

• Inferred context through interpretation and reasoning

• Open domain spoken dialog systems

• Open domain spoken/natural language understanding and generation

• Open domain semantic interpretation

• Open domain dialog management (large-scale belief state/policy)

• Conversational Interactions

• Multi-modal inputs in situated open domains (speech/text + gesture, touch, eye gaze)

• Multi-human situated interactions


Larry Heck, Microsoft Research, larry [at]

Dilek Hakkani-Tür, Microsoft Research, dilek [at]

Gokhan Tur, Microsoft Research, gokhan [at]

Steve Young, Cambridge University, sjy [at]

Phase Importance in Speech Processing Applications

In the past decades, the amplitude of speech spectrum is considered to be the most important feature in

different speech processing applications and phase of the speech signal has received less

attention. Recently, several findings justify the phase importance in speech and audio processing communities.

The importance of phase estimation along with amplitude estimation in speech enhancement,

complementary phase-based features in speech and speaker recognition and phase-aware acoustic

modeling of environment are the most prominent

reported works scattered in different communities of speech and audio processing. These examples suggest

that incorporating the phase information can push the limits of state-of-the-art phase-independent solutions

employed for long in different aspects of audio and speech signal processing. This Special Session aims

to explore the recent advances and methodologies to exploit the knowledge of signal phase information in different

aspects of speech processing. Without a dedicated effort to bring researchers from different communities,

a quick advance in investigation towards the phase usefulness in speech processing applications

is difficult to achieve. Therefore, as the first step in this direction, we aim to promote the 'phase-aware

speech and audio signal processing' to form a community of researchers to organize the next steps.

Our initiative is to unify these efforts to better understand the pros and cons of using phase and the degree

of feasibility for phase estimation/enhancement in different areas of speech processing including: speech

enhancement, speech separation, speech quality estimation, speech and speaker recognition,

voice transformation and speech analysis and synthesis. The goal is to promote the importance of

the phase-based signal processing and studying its importance and sharing interesting findings from different

speech processing applications.


Pejman Mowlaee, Graz University of Technology, pejman.mowlaee [at]

Rahim Saeidi, University of Eastern Finland, rahim.saeidi [at]

Yannis Styilianou, Toshiba Labs Cambridge UK / University of Crete, yannis [at]

Speaker Comparison for Forensic and Investigative Applications

In speaker comparison, speech/voice samples are compared by humans and/or machines

for use in investigation or in court to address questions that are of interest to the legal system.

Speaker comparison is a high-stakes application that can change people’s lives and it demands the best

that science has to offer; however, methods, processes, and practices vary widely.

These variations are not necessarily for the better and though recognized, are not generally appreciated

and acted upon. Methods, processes, and practices grounded in science are critical for the proper application

(and non-application) of speaker comparison to a variety of international investigative and forensic applications.

This special session will contribute to scientific progress through 1) understanding speaker comparison

for investigative and forensic application (e.g., describe what is currently being done and critically

analyze performance and lessons learned); 2) improving speaker comparison for investigative and forensic

applications (e.g., propose new approaches/techniques, understand the limitations, and identify challenges

and opportunities); 3) improving communications between communities of researchers, legal scholars,

and practitioners internationally (e.g., directly address some central legal, policy, and societal questions

such as allowing speaker comparisons in court, requirements for expert witnesses, and requirements for specific

automatic or human-based methods to be considered scientific); 4) using best practices (e.g., reduction of bias

and presentation of evidence); 5) developing a roadmap for progress in this session and future sessions; and 6)

producing a documented contribution to the field. Some of these objectives will need multiple sessions

to fully achieve and some are complicated due to differing legal systems and cultures.

This special session builds on previous successful special sessions and tutorials in forensic applications

of speaker comparison at INTERSPEECH beginning in 2003. Wide international participation is planned,

including researchers from the ISCA SIGs for the Association Francophone de la Communication Parlée (AFCP)

and the Speaker and Language Characterization (SpLC).


Joseph P. Campbell, PhD, MIT Lincoln Laboratory, jpc [at]

Jean-François Bonastre, l'Université d'Avignon, jean-francois.bonastre [at]

Text-dependent for Short-duration Speaker Verification

In recent years, speaker verification engines have reached maturity and have been deployed in

commercial applications. Ergonomics of such applications is especially demanding and imposes

a drastic limitation in terms of speech duration during authentication. A well known tactic to address

the problem of lack of data, due to short duration, is using text-dependency. However, recent breakthroughs

achieved in the context of text-independent speaker verification in terms of accuracy and robustness

do not benefit text-dependent applications. Indeed, large development data required by the recent

approaches is not available in the text-dependent context. The purpose of this special session is

to gather the research efforts from both academia and industry toward a common goal of establishing

a new baseline and explore new directions for text-dependent speaker verification.

The focus of the session is on robustness with respect to duration and modeling of lexical information.

To support the development and evaluation of text-dependent speaker verification technologies,

the Institute for Infocomm Research (I2R) has recently released the RSR2015 database,

including 150 hours of data recorded from 300 speakers. The papers submitted to the special

session are encouraged, but not limited, to provide results based on the RSR2015 database

in order to enable comparison of algorithms and methods. For this purpose, the organizers strongly

encourage the participants to report performance on the protocol delivered with the database

in terms of EER and minimum cost (in the sense of NIST 2008 Speaker Recognition evaluation).

To get the database, please contact the organizers.

Further details are available at:


Anthony LARCHER (alarcher [at] Institute for Infocomm Research

Hagai ARONOWITZ (hagaia [at] IBM Research – Haifa

Kong Aik LEE (kalee [at] Institute for Infocomm Research

Patrick KENNY (patrick.kenny [at] CRIM – Montréal

Tutorial Dialogues and Spoken Dialogue Systems

The growing interest in educational applications that use spoken interaction and dialogue technology has boosted

research and development of interactive tutorial systems, and over the recent years, advances have been achieved

in both spoken dialogue community and education research community, with sophisticated speech and multi-modal

technology which allows functionally suitable and reasonably robust applications to be built.

The special session combines spoken dialogue research, interaction modeling, and educational applications,

and brings together the two INTERSPEECH SIG communities: SLaTE and SIGdial. The session focuses

on methods, problems and challenges that are shared by both communities, such as sophistication

of speech processing and dialogue management for educational interaction, integration of the models

with theories of emotion, rapport, and mutual understanding, as well as application of the techniques

to novel learning environments, robot interaction, etc. The session aims to survey issues related

to the processing of spoken language in various learning situations, modeling of the teacher-student

interaction in MOOC-like environments, as well as evaluating tutorial dialogue systems from

the point of view of natural interaction, technological robustness, and learning outcome.

The session encourages interdisciplinary research and submissions related to the special focus

of the conference, 'Celebrating the Diversity of Spoken Languages'.

For further information click


Maxine Eskenazi, max+ [at]

Kristiina Jokinen, kristiina.jokinen [at]

Diane Litman, litman [at]

Martin Russel, M.J.RUSSELL [at]

Visual Speech Decoding

Speech perception is a bi-modal process that takes into account both the acoustic (what we hear)

and visual (what we see) speech information. It has been widely acknowledged that visual clues play

a critical role in automatic speech recognition (ASR) especially when audio is corrupted by,

for example, background noise or voices from untargeted speakers, or even inaccessible.

Decoding the visual speech is utterly important for ASR technologies to be widely implemented

to realize truly natural human-computer interactions. Despite the advances in acoustic ASR,

visual speech decoding remains a challenging problem.

The special session aims to attract more effort to tackle this important problem. In particular,

we would like to encourage researchers to focus on some critical questions in the area.

We propose four questions as the initiative as follows:

1. How to deal with the speaker dependency in visual speech data?

2. How to cope with the head-pose variation?

3. How to encode temporal information in visual features?

4. How to automatically adapt the fusion rule when the quality of the two individual (audio and visual)

modalities varies?

Researchers and participants are encouraged to raise more questions related to visual speech decoding.

We expect the session to draw a wide range of attention from both the speech recognition and machine vision

communities to the problem of visual speech decoding.


Ziheng Zhou, University of Oulu, ziheng.zhou [at]

Matti Pietikäinen, University of Oulu, matti.pietikainen [at]

Guoying Zhao, University of Oulu, gyzhao [at]


3-1-18(2015-09-06) Call for Satellite Workshops of INTERSPEECH 2015, Dresden, Germany
**** Call for Satellite Workshops **** 
INTERSPEECH 2015 will be held in the beautiful city of Dresden, Germany, on September 6-10, 2015
The theme is 'Speech beyond Speech - Towards a Better Understanding of the Most Important 
Biosignal'. The Organizing Committee of INTERSPEECH 2015 is now inviting proposals for 
satellite workshops, which will be held in proximity to the main conference. 
The Organizing Committee will work to facilitate the organization of such satellite workshops, 
to stimulate discussion in research areas related to speech and language, at locations in Central 
Europe, and around the same time as INTERSPEECH. We are particularly looking forward to 
proposals from neighboring countries. If you are interested in organizing a satellite workshop, 
or would like a planned event to be listed as an official satellite event, please contact the organizers
 or the Satellite Workshop Chair at The Satellite Workshop coordinator along 
with the INTERSPEECH team will help to connect (potential) workshop organizers with local 
contacts in Germany, if needed, and will try to be helpful with logistics such as payment, publicity,
 and coordination with ISCA or other events. Proposals should include:
 * workshop name and acronym 
* organizers' name and contact info 
* website (if already known) 
* date and proposed location of the workshop 
* estimated number of participants 
* a short description of the motivation for the workshop 
* an outline of the program and invited speakers 
* a description of the submission process (e.g. deadlines, target acceptance rate) 
* a list of the scientific committee members 
Proposals for satellite workshops should be submitted by email to
 by August 31st, 2014 We strongly recommend that organizers also apply for
 ISCA approval/ sponsorship, which will greatly facilitate acceptance as an INTERSPEECH satellite 
event. We plan to notify proposers no later than October 30, 2014. If you have any questions about 
whether a potential event would be a good candidate for an INTERSPEECH 2015 satellite workshop 
feel free to contact the INTERSPEECH 2015 Satellite Workshops Chair. 
Florian Metze 
Satellite Workshops Chair



3-1-19(2015-09-06) INTERSPEECH 2015 Dresden RFA

Interspeech 2015


September 6-10, 2015, Dresden, Germany



Speech Beyond Speech: Towards a Better Understanding of the Most Important Biosignal



Speech is the most important biosignal humans can produce and perceive. It is the most common means of human-human communication, and therefore research and development in speech and language are not only paramount for understanding humans, but also to facilitate human-machine interaction. Still, not all characteristics of speech are fully understood, and even fewer are used for developing successful speech and language processing applications. Speech can exploit its full potential only if we consider the characteristics which are beyond the traditional (and still important) linguistic content. These characteristics include other biosignals that are directly accessible to human perception, such as muscle and brain activity, as well as articulatory gestures.



will therefore be organized around the topic “Speech beyond Speech: Towards a Better Understanding of the Most Important Biosignal”. Our conviction is that spoken language processing can make a substantial leap if it caters for the full information which is available in the speech signal. By opening our prestigious conference to researchers in other biosignal communities, we expect that substantial advances can be made discussing ideas and approaches across discipline and community boundaries.







The following people organize INTERSPEECH 2015:


  • General Chair: Sebastian Möller, Telekom Innovation Laboratories, Technische Universität Berlin
  • General Co-Chair & International Outreach: Hermann Ney, Chair of Computer Science 6, RWTH Aachen University
  • Technical Program: Bernd Möbius, Dept. of Computational Linguistics and Phonetics, Universität des Saarlandes; Elmar Nöth, Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg
  • Local Organization: Rüdiger Hoffmann, Senior Professor, Technische Universität Dresden; Ercan Altinsoy and Ute Jekosch, Chair for Communication Acoustics, Technische Universität Dresden
  • Plenaries: Gerhard Rigoll, Institute for Human-Machine Communication, Technische Universität München
  • Special Sessions & Challenges: Anton Batliner, Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg; Björn Schuller, Imperial College London & Technische Universität München
  • Tutorials: Alexander Raake, Telekom Innovation Laboratories, Technische Universität Berlin
  • Satellite Workshops: Florian Metze, Carnegie Mellon University, Pittsburgh
  • Industry Liaison: Jimmy Kunzmann, EML European Media Laboratory GmbH, Heidelberg
  • Sponsoring: Tim Fingscheidt, Institute for Communications Technology, Technische Universität Braunschweig; Claudia Pohlink, Telekom Innovation Laboratories, Deutsche Telekom AG, Berlin
  • Special Events: David Sündermann, Duale Hochschule Baden-Württemberg, Stuttgart
  • Show and Tell: Georg Stemmer, Intel, München
  • Social Events: Petra Wagner, Phonetics and Phonology Workgroup, Universität Bielefeld
  • Exhibits: Reinhold Häb-Umbach, Universität Paderborn
  • Finance: Volker Steinbiss, RWTH Aachen University and Accipio Projects GmbH, Aachen
  • Community Outreach: Norbert Reithinger, DFKI Projektbüro Berlin
  • Publicity: Oliver Jokisch, Hochschule für Telekommunikation Leipzig
  • Publications: Stefan Steidl, Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg
  • Students Affairs: Benjamin Weiss, Telekom Innovation Laboratories, Technische Universität Berlin
  • Web & Tools: Tim Polzehl, Telekom Innovation Laboratories, Technische Universität Berlin
  • Grants: Michael Wagner, University of Canberra
  • PCO: Lisa Hertel, TUBS GmbH TU Berlin ScienceMarketing


The event will be staged in the recently built Maritim International Congress Center (ICD) in Dresden, Germany. As the capital of Saxony, an up-and-coming region located in the former eastern part of Germany, Dresden combines glorious and painful history with a strong dedication to future and technology. It is located in the heart of Europe, easily reached via two airports, and will offer a great deal of history and culture to INTERSPEECH 2015 delegates. Guests are well catered for in a variety of hotels of different standards and price ranges, making INTERSPEECH 2015 an exciting as well as an affordable event.



Prof. Dr.-Ing. Sebastian Möller, Quality and Usability Lab, Telekom Innovation Laboratories, TU Berlin

Sekr. TEL-18, Ernst-Reuter-Platz 7, D-10587 Berlin, Germany





3-1-20(2016) INTERSPEECH 2016, San Francisco, CA, USA

Interspeech 2016 will take place

from September 8-12 2016 in San Francisco, CA, USA

General Chair is Nelson Morgan.



 Organisation  Events   Membership   Help 
 > Board  > Interspeech  > Join - renew  > Sitemap
 > Legal documents  > Workshops  > Membership directory  > Contact
 > Logos      > FAQ
       > Privacy policy

© Copyright 2024 - ISCA International Speech Communication Association - All right reserved.

Powered by ISCA