(2017-02-13) CDD POST-DOCTORANT 18 mois *Analyse multimodale de contenus audiovisuels*
*CDD POST-DOCTORANT 18 mois* *Analyse multimodale de contenus audiovisuels*
L?équipe LINKMEDIA (IRISA & Inria Rennes) travaille au développement des futurs technologies permettant la description et l?accès aux contenus multimédias par le biais de leur analyse. Les domaines de compétence de l?équipe sont la vision par ordinateur, le traitement de la parole et du langage, le traitement des contenus audio, la recherche d?information et la fouille de données. En particulier, l?équipe participe au projet FUI NexGenTV portant sur l?analyse et l?enrichissement de contenus télévisés. La télévision évolue de l?écran du téléviseur vers des applications multi-écrans où le spectateur regarde la télévision tout en explorant le web, cherchant des compléments d?informations ou réagissant sur les réseaux sociaux. Dans ce contexte, NexGenTV cherche à apporter des solutions d?édition de contenus enrichis multi-écrans par le bais de fonctionnalités telles la détection de temps forts, l?enrichissement de programmes par des informations complémentaires et, plus généralement, l?optimisation de l?expérience utilisateur en favorisant l?interaction adaptée aux attentes de l?utilisateur. Au sein du projet, l?IRISA s?intéresse à l?analyse des contenus audiovisuels, de la parole et des réseaux sociaux.
Dans ce contexte, nous souhaitons recruter un chercheur post-doctorant spécialisé dans l?analyse de contenus audiovisuels pour développer, étudier et évaluer des approches innovantes relatives à l?analyse des personnes au sein des contenus télévisés. On cherchera notamment à concevoir des approches multimodales (voix+visage) permettant aussi bien la détection de personnes connues que la mise en relation de vidéos d?un même intervenant. Une première piste de travail s?appuie sur des travaux récents de l?équipe en apprentissage de représentations multimodales à l?aide de réseaux neuronaux. On pourra également étudier l?usage de ces derniers pour la représentation et la comparaison des voix. Dans un second temps, on s?intéressera à l?exploitation de tels modèles pour enrichir un contenu live avec des extraits de documents archivés, combinant identification des intervenants et pertinence sémantique.
Les recherches envisagées seront menées dans l?équipe LINKMEDIA de l?IRISA (Rennes, France), en collaboration étroite avec les partenaires du projet NexGenTV, notamment avec EURECOM.
Le candidat devra posséder une thèse dans un domaine proche du sujet de recherche, de préférence dans l?un des domaines suivants : modélisation multimodale, traitement automatique de la parole, reconnaissance du locuteur, vision par ordinateur. On attend également du candidat qu?il renforce la compétence de l?équipe en apprentissage neuronal appliqué à l?analyse des contenus multimédia.
Pour candidater, merci d?adresser un CV accompagné d?une lettre de motivation.
Employeur : Centre National de la Recherche Scientifique Lieu d?exercice : IRISA, Rennes Contrat : CDD 18 mois, dès que possible à partir de mars 2017 Rémunération : 2 815? mensuels bruts Contact : Guillaume Gravier (prenom.nom@irisa.fr)
(2017-02-20) Several possitions of Research Engineers at Audio Analytic Labs, Cambridge, UK
Audio Analytic Labs, the research division of Audio Analytic Ltd, has several Research Engineer positions currently open in the field of Automatic Sound Recognition.
These could be of interest to your PhD students or Post-Docs finishing their contracts in your teams and looking to follow up with an industrial position.
The complete job specification is copied below.
We are also open to answering questions from people interested in our company but not yet available for employment.
More generally, we are open to finding concrete and mutually beneficial ways to collaborate with academic partners on research projects, either through joint projects supported by specific funding, or via secondments and internships.
For more information about our company, please visit the company?s website on http://www.audioanalytic.com/ , or feel free to contact me directly.
I would be very grateful if you could propagate the attached job offer to your institutions? career services, or if you could forward it directly to people who you think may be directly interested in applying.
Hoping this will be useful, and of interest to your alumni.
Many thanks, and best regards,
- Sacha K.
Director of AALabs
AudioAnalytic Ltd.
INDUSTRIAL POSITION OPEN:
*Full Time Audio Analytics Research Engineer*
Location: Cambridge, Cambridgeshire, United Kingdom
Full-time, immediate start.
Audio Analytic Ltd. is leading the world of acoustically connected things. Our unique software is used by smart home companies the world over to make devices aware of sounds around them. If a smoke alarm goes off or a glass panel is broken by intruders while no-one is at home, our software will immediately recognise the sound and tell the device to alert the home owner and the smart home so they can both take appropriate protective action. We give smart home owners sound peace of mind. More information is available on:
We are looking for people who thrive as part of a dedicated and innovative team, love tough challenges, and are passionate about audio/sound, DSP and Machine Learning.
Responsibilities
As part of our R&D team, you will contribute to researching and evaluating new algorithms to push the limits of our unique sound recognition system. Responsibilities include developing new algorithms in house, identifying and reporting on state of the art methods, and evaluating both types of solutions on large scale field data sets.
Technical Skills
Must have either a Master?s degree with 2 years industrial experience or a PhD, in one of the following topics: Digital Signal Processing of Audio Signals, Machine Learning applied to Audio Signals, Automatic Speech/Speaker Recognition, Music Information Retrieval, Acoustic Events Detection, Statistical Speech Synthesis, Thematic Indexing of Audio tracks (e.g., Speaker Diarization, Acoustic Segmentation of Video Documents etc.).
Experience as a post-doc research engineer, either academic or industrial, will be a significant plus.
Required:
Demonstrable skills in Digital Signal Processing and/or Machine Learning applied to Audio Signals.
Demonstrable experience dealing with at least one type of Machine Learning algorithm (e.g., Deep Neural Networks, Hidden Markov Models, Support Vector Machines, Decision Trees etc.) applied to the processing of Audio Signals.
Scripting and algorithm prototyping: Python, bash.
Programming: C/C++ coding and code optimisation. CUDA/GPU programming a plus.
Development under Linux/Unix mandatory, Windows optional.
Desirable:
Hardware design knowledge a plus but not a requirement.
Demonstrable interest in porting DSP/Machine Learning algorithms to either embedded platforms or high performance computing platforms a plus but not a requirement.
General Skills
Ability to deliver on research and evaluation methodology.
Good communication skills.
Excellent problem-solving skills.
Track record of academic publications a plus but not a requirement.
Enjoy working as a member of a team and using their own initiative.
Self-confident and highly motivated.
Ability to deal confidently with a variety of people at all levels.
Able to manage own workload and meet deadlines.
Good organisational skills.
Good standard of written and spoken English.
Remuneration
This is a great opportunity to join a successful company with a huge potential for growth. The successful candidate will be compensated with an attractive package appropriate to qualifications and experience, to include a competitive salary and stock options.
How to Apply
To apply for this vacancy, please send a covering letter and copy of a recent CV to jobs@audioanalytic.com, with reference AA-RES-ENG-2016 in the email title.
Please note that it is company policy not to accept job applications from recruitment consultants.
(2017-02-21) Acting Assistant Professor, Department of Linguistics, University of Washington, WA, USA
Acting Assistant Professor, Department of Linguistics, University of Washington, Washington, USA, associated with the professional MS program and Ph.D. track in Computational Linguistics.
Fluent.ai is a startup based in Montreal, Canada. We are working on new deep learning and related techniques to enable acoustic-only speech recognition. By associating speech to intent without requiring a speech-to-text translation, Fluent.ai opens a wide variety of new applications and provides higher accuracy and more robust performance compared to existing methods. We are looking to expand our technology and research teams and are inviting applications for various permanent and internship based roles. Joining Fluent.ai provides you an opportunity to be an early team member leading work on an exciting, disruptive technology poised for rapid growth. The technology has already been validated by many academic experts as well as industrial customers in diverse sectors. Now we are looking for the right people to share our vision and hustle to achieve execution excellence in select sectors. You will be joining a diverse, dedicated, smart and fun team. We work hard, we don?t always agree, but we always laugh out loud and we always move forward together. What we offer: We offer a great working environment and a competitive mix of salary and options. We are keen to interact with talented people and will get back to the selected candidates quickly. We are an equal opportunity employer and value diversity at our company. We do not discriminate based on origin, religion, gender, age, sexual orientation, or disability. We are looking for both permanent full-time employees as well as interns. Please link this page: http://www.fluent.ai/careers/#toggle-id-3 Let me know if you have any questions, and I will be happy to answer those. About Fluent.ai Fluent.ai is a startup based in Montreal, Canada. We are working on new deep learning and related techniques to enable acoustic-only speech recognition. By associating speech to intent without requiring a speech-to-text translation, Fluent.ai opens a wide variety of new applications and provides higher accuracy and more robust performance compared to existing methods. We are looking to expand our technology and research teams and are inviting applications for various permanent and internship based roles. Joining Fluent.ai provides you an opportunity to be an early team member leading work on an exciting, disruptive technology poised for rapid growth. The technology has already been validated by many academic experts as well as industrial customers in diverse sectors. Now we are looking for the right people to share our vision and hustle to achieve execution excellence in select sectors. You will be joining a diverse, dedicated, smart and fun team. We work hard, we don?t always agree, but we always laugh out loud and we always move forward together. What we offer: We offer a great working environment and a competitive mix of salary and options. We are keen to interact with talented people and will get back to the selected candidates quickly. We are an equal opportunity employer and value diversity at our company. We do not discriminate based on origin, religion, gender, age, sexual orientation, or disability.
(2017-02-22) Language modeling scientist at Siri team at Apple
Title: Language Modeling Scientist – Siri Speech team at Apple
Job Summary
Play a part in the next revolution in human-computer interaction. Contribute to a product that is redefining mobile computing. Create groundbreaking technology for large scale systems, spoken language, big data, and artificial intelligence. And work with the people who created the intelligent assistant that helps millions of people get things done — just by asking. Join the Siri Speech team at Apple.
The Siri team is looking for exceptionally skilled and creative Scientists and Engineers eager to get involved in hands-on work improving the Siri experience.
Key Qualifications
Experience building, testing, and tuning language models for ASR
Ability to implement experiments using scripting languages (Python, Perl, bash) and tools written in C/C++
Experience working with standard speech recognition toolkits (such as HTK, Attila, Kaldi, SRILM, OpenFST or equivalent proprietary systems)
Large scale data analysis experience using distributed clusters (e.g. MapReduce, Spark)
Description
The speech team is seeking a research scientist to participate in the language modeling effort for Siri. In order to estimate language model probabilities, you will make use of very large amounts of training text drawn from diverse sources. You will be part of a group that has responsibility for the entire domain of language modeling in multiple languages including, among other things, text processing, data selection, language model adaptation, neural network modeling, improving language model training infrastructure, experimenting with new types of language models etc.
Education
PhD or Masters in Computer Science or related field
3+ years of experience in language modeling for ASR
The University of Eastern Finland, UEF, is one of the largest multidisciplinary universities in Finland. We offer education in nearly one hundred major subjects, and are home to approximately 15,000 students and 2,800 members of staff. We operate on three campuses in Joensuu, Kuopio and Savonlinna. In international rankings, we are ranked among the leading 300 universities in the world.
The Faculty of Science and Forestry operates on the Kuopio and Joensuu campuses of the University of Eastern Finland. The mission of the faculty is to carry out internationally recognised scientific research and to offer research-education in the fields of natural sciences and forest sciences. The faculty invests in all of the strategic research areas of the university. The faculty?s environments for research and learning are international, modern and multidisciplinary. The faculty has approximately 3,800 Bachelor?s and Master?s degree students and some 490 postgraduate students. The number of staff amounts to 560. http://www.uef.fi/en/lumet/etusivu
We are now inviting applications for
a Postdoctoral Researcher (Speech/Audio Processing), School of Computing, Joensuu Campus
The Machine Learning research group of the School of Computing at the University of Eastern Finland (http://www.uef.fi/en/web/cs) is looking for a highly motivated researcher to work in the group.
The current research topics in the group include speaker and language recognition, voice conversion, spoofing and countermeasures for speaker recognition, robust feature extraction, and analysis of environmental sounds. Prior experience in these topics is a plus, though we invite candidates widely from general speech/audio/language processing, machine learning or signal processing background. We expect the new Postdoctoral Researcher to bring in complementary skills and expertise.
The recruited Postdoctoral Researcher will take a major role in advancing research in one of the above-listed (or closely related) topics. He or she will also have a significant role in the supervision of students and certain administrative duties, and he or she will work closely with Associate Professor Kinnunen and the other members of the group. The position is strongly research-focused.
The School of Computing, located in Joensuu Science Park, provides modern research facilities with access to high-performance computing services. Our research group hosted the Odyssey 2014 conference (http://cs.uef.fi/odyssey2014/), is a partner in the ongoing H2020 funded OCTAVE project (https://www.octave-project.eu/) focused on voice biometrics, is a co-founder of the Automatic Speaker Verification and Countermeasures (ASVspoof) challenge series (http://www.spoofingchallenge.org/) and has hosted international summer schools. We take actively part in international benchmarking and other collaboration activities. We follow a multidisciplinary research perspective that targets at understanding the speech signal, as well as applying the acquired knowledge to new application areas.
A person to be appointed as a postdoctoral researcher shall hold a suitable doctoral degree that has been awarded less than five years ago. The doctoral degree should be in spoken language technology, electrical engineering, computer science, machine learning or a closely related field. He/she should be comfortable with Unix/Linux, Matlab/Octave and/or Python, processing of large datasets and with strong hands-on experience and creative out-of-the-box problem solving attitude.
The position will be filled from May 1, 2017 until December 31, 2018. The continuation of the position will be agreed separately.
The positions of postdoctoral researcher shall always be filled for a fixed term (UEF University Regulations 31 §).
The salary of the position is determined in accordance with the salary system of Finnish universities and is based on level 5 of the job requirement level chart for teaching and research staff (?2.865,30/ month). In addition to the job requirement component, the salary includes a personal performance component, which may be a maximum of 46.3% of the job requirement component.
For further information on the position, please contact: Associate Professor Tomi Kinnunen, email: tkinnu@cs.uef.fi, tel. +358 50 442 2647. For further information on the application procedure, please contact: Executive Head of Administration Arja Hirvonen, tel. +358 44 716 3422, email: arja.hirvonen@uef.fi.
A probationary period is applied to all new members of the staff.
The electronic application should contain the following appendices:
- a cover letter indicating the position to be applied for and a free-worded motivation letter - a résumé or CV - a list of publications - copies of the applicant's academic degree certificates/ diplomas, and copies of certificates / diplomas relating to the applicant?s language proficiency, if not indicated in the academic degree certificates/diplomas - the names and contact information of at least two referees
The application needs to be submitted no later than March 24, 2017 (by 24:00 EET) by using the electronic application form:
The job ad and the application form can also be located under http://www.uef.fi/en/uef/en-open-positions (seek for the position 'Postdoctoral Researcher (Speech/Audio Processing)').
(2017-02-28) MCF en informatique pour les Sciences Humaines, Sorbonne, Paris, France
Un poste de MCF en informatique pour les Sciences Humaines, notamment en traitement automatique du langage et/ou de la parole, est ouvert à l'Université Paris Sorbonne (www.paris-sorbonne.fr/IMG/pdf/27-7_mcf_766.pdf). Le candidat enseignera l'Informatique dans les différentes formations de licence et de master du département d'Informatique, Mathématiques et de Linguistique appliquées de l'UFR de Sociologie et d'Informatique pour les Sciences Humaines. Il devra s'inscrire dans un ou plusieurs axes de l'équipe de linguistique computationnelle (www.stih.paris-sorbonne.fr/) : (1) Sémantiques, connaissances et corpus (2) Paralinguistique, cognition et physiologie.
Personne à contacter : Claude.Montacie@Paris-Sorbonne.fr
(2017-03-14) PhD and postdocs positions at INRIA/Nancy France
Our team has several openings for PhD students and postdocs in the fields of deep learning based: - speech enhancement - speech recognition - environmental sound analysis
(2017-03-18) Fully-funded PhD Positions in Automatic Emotion Recognition at SUNY, Albany, NY, USA
Fully-funded PhD Positions in Automatic Emotion Recognition at SUNY Application deadline: 22 March 2017 (**see below for more information**)
We have several PhD research assistantship positions available at the State University of New York, Albany. We are seeking highly creative and motivated applicants with a keen interest in doing research in human-centered technology, affective computing, and automatic emotion recognition using machine learning and multimodal signal processing techniques.
Requirements: - A bachelor's degree in a relevant field (Electrical and Computer Engineering, Computer Science, Statistics, or related) - Solid background in computer programming - Proficiency in spoken and written English - (Preferred) Knowledge in the following technologies: MATLAB, Python, Java, Perl, C++, Unity - (Preferred) Previous coursework and/or practical experience in machine learning - (Preferred) Solid background in mathematics and/or statistics Interest in one of the following areas: - Human-Centered and Affective Computing, Computational Human Behavior Analysis - Machine Learning, Statistics, Applied Mathematics - Speech Processing, Computer Vision We expect: - Keen interest in top level conference and journal publications - Self-organized, team worker, with good communication skills We offer: - You will work at one of the leading U.S. Universities and have the opportunity to work towards your PhD in a group of excellent scientists - Tuition, stipend, and fringe benefits - You will get financial support to attend and present at top level international conferences - Visas will be fully funded for international students
To apply, please send an email to Prof. Yelin Kim (yelinkim@albany.edu) including a CV and a research statement (max. 2 pages) by March 22, 2017. We have rolling admissions policies, so please apply as early as possible. Please give your email the subject “SUNY PhD Research Assistantship in Automatic Emotion Recognition.'
Please liberally forward and share to possibly interested candidates or people that might know suitable candidates.
(2017-03-20) Ph D position at IRISA Rennes, France
The Expression team of IRISA is recruiting a PhD candidate in computer science on the subject 'Universal speech synthesis through embeddings of massive heterogeneous data'. This work focuses on the following domains:
- Text-to-speech
- Deep learning
- High-dimensional indexing.
Details are given here: http://www.irisa.fr/en/offres-theses/universal-speech-synthesis-through-embeddings-massive-heterogeneous-data .
Application deadline: Monday, 3 April 2017.
Application process:
- CV
- Transcript of M.Sc. marks/grades
- to gwenole.lecorve@irisa.fr, damien.lolive@irisa.fr, laurent.amsaleg@irisa.fr .
(2017-03-25) Offre de thèse en Systèmes d'interaction vocale , LIA, Avignon France
***** Offre de thèse en Systèmes d?interaction vocale ***** au LIA/CERI Univ. Avignon Prof. F. Lefèvre et B. Jabaian
Améliorer l'interaction vocale avec le monde numérique et la conception de nouveaux services de dialogue homme-machine sont des défis essentiels pour un passage total vers une société numérique. Parmi les activités de recherche en intelligence artificielle portant sur les interactions vocales, plusieurs questions importantes sont encore mal examinées et peuvent faire l?objet de différentes études. Le LIA traite de multiples aspects liés à l?interaction vocale et cherche à travers cette thèse à approfondir la recherche dans l?une des ces grandes problématiques parmi :
** Le dialogue argumentatif ** pour rendre les systèmes artificiels capables d'apprendre à partir des données, deux hypothèses fortes sont généralement faites : (1) la stationnarité du système : on suppose que l'environnement de la machine ne changera pas avec le temps. (2) l'interdépendance entre la collecte des données et le processus d'apprentissage : cela implique que l'utilisateur ne modifie pas son comportement dans le temps alors que ce dernier a tendance à adapter son comportement en fonction de la réaction de la machine. Il est clair que ce comportement n'aide pas un système d'apprentissage artificiel à trouver l'équilibre lui permettant de satisfaire au mieux les attentes de l'utilisateur.
Les interfaces vocales actuelles, basées sur des processus de décision markovien partiellement observables, doivent évoluer vers une nouvelle génération de systèmes interactifs, capables d'apprendre dynamiquement à partir d'interactions sur le long terme, tout en tenant compte que le comportement des humains est variable, étant eux-mêmes des systèmes adaptatifs. En effet, les humains apprennent également de leurs interactions avec un système et changent leur comportement au cours du temps. Un tel système sera capable de discuter avec l?humain et argumenter pour défendre ses choix.
** L?agent dialoguant autoritaire ** L'intelligence artificielle est généralement vue à travers sa soumission aux désirs/volontés de l'humain, il existe toutefois des situations où artificiellement doter la machine d'une dimension autoritaire peut être pertinent (games et serious games principalement, mais aussi simulation de contrôle...). Des mécanismes concrets permettant de développer un agent autoritaire (dans l'objectif d'imposer son point de vue à l'utilisateur) seront étudiés et mis en oeuvre en pratique pour permettre leur évaluation complète.
** La réalité virtuelle pour la simulation d'agents dialoguant ** Une autre piste de recherche concerne les possibilités offertes par la réalité virtuelle pour permettre l'apprentissage d'agent vocaux dialoguant. L'objectif initial est d'offrir un cadre unifié pour le développement en conditions d'utilisation de systèmes de dialogue situés par le biais de simulations en réalité virtuelle des environnements envisagés, éliminant ainsi la nécessité de les recréer. A terme l'approche permettra aussi de développer des systèmes de dialogue pour les applications de réalité virtuelle elle-mêmes. Le travail implique donc des compétences dans les deux domaines de la réalité virtuelle et du traitement automatique du langage.
Le candidat doit avoir un master en informatique avec une composante sur les méthodes d'apprentissage automatique et/ou sur l?ingénierie de la langue. La bourse de thèse fera l?objet d?un concours au sein de l?Ecole Doctorale 536 de l?université d?Avignon, avec une audition du candidat retenu par les encadrants de thèse.
Pour postuler merci d?envoyer un mail avant le 30 avril 2017 à Fabrice Lefèvre (fabrice.lefevre@univ-avignon.fr) et Bassam Jabaian (bassam.jabaian@univ-avignon.fr) incluant : votre CV, une lettre de motivation avec votre positionnement sur les propositions d?études ci-dessus, d?éventuelles lettres de recommandation et vos relevés de notes.
(2017-03-28) Research Scientist, Spoken and Multimodal Dialog Systems, ETS, S.Francisco, CA, USA
Open Rank Research Scientist, Spoken and Multimodal Dialog Systems
ETS (Educational Testing Service) is a global not for profit organization whose mission is to advance quality and equity in education. With more than 3,400 global employees, we develop, administer and score more than 50 million tests annually in more than 180 countries.
Our San Francisco Research and Development division is seeking a Research Scientist for our Dialog, Multimodal, and Speech (DIAMONDS) research center. The center’s main focus is on foundational research as well as on development of new capabilities to automatically score spoken, interactive, and multimodal test responses in conversational settings in a wide range of ETS test programs, promote learning and other educational areas. This is an excellent opportunity to be part of a world-renowned research and development team and have a significant impact on existing and next generation spoken and multimodal dialog systems and their application to assessment and other areas in education.
Primary responsibilities include:
Developing and collaborating on interdisciplinary projects that aim to transfer techniques to a new context or scientific field. Successful candidates are self-motivated and self-driven, and have a strong interest in emerging conversational technology that can contribute to education in assessment and instructional settings.
Providing scientific and technical skills to conceptualize, design, obtain support for, conduct, and manage new research projects, grants, or parts of existing projects.
Generating or contributing to new or modified methods that support research on and development of spoken and multimodal dialog systems and related technologies relevant in assessment and instructional settings.
Designing and conducting scientific studies and functioning as an expert in the major facets of the projects: responding as a subject matter expert in presenting the results of acquired knowledge and experience.
Developing or assisting in developing proposals for external and internal research grants and obtain financial support for new or continuing research activities. Prepare initial and final proposal and project budgets.
Participating in dissemination activities through the publications of research papers in peer-reviewed journals and in the ETS Research Report series, the issuing of progress and technical reports, the presentation of seminars at major conferences and at ETS, or the use of other appropriate communication vehicles, including patents, books and chapters, that impact practice in the field or at ETS.
Depending on experience this position is open to entry level candidates as well as mid-level and senior level professionals.
REQUIREMENTS FOR A JUNIOR LEVEL POSITION
A Doctorate in computer science, linguistics, cognitive psychology or a related field is required. One year of research experience is required, in education is desirable. Experience can be gained through doctoral studies. Candidates should be very skilled in programming and be able to work effectively as a research team member.
REQUIREMENTS FOR A MID-LEVEL POSITION
A Doctorate in computer science, linguistics, cognitive psychology, or a related field is required. Research experience in education is desirable. Candidates should be very skilled in programming and be able to work effectively as a research team member. Three years of progressively independent substantive research in the area of computer science, linguistics, cognitive psychology, or education are required.
REQUIREMENTS FOR A SENIOR-LEVEL POSITION
A Doctorate in computer science, linguistics, cognitive psychology, or a related field is required. Research experience in education is desirable. Candidates should be very skilled in programming and be able to work effectively as a research team member. Eight years of progressively independent substantive research in the area of computer science, linguistics, cognitive psychology, or education are required.
1. Embedding enhancement information in the speech signal
Speech becomes harder to understand in the presence of noise and other distortions, such as telephone channels. This is especially true for people with a hearing impairment. It is difficult to enhance the intelligibility of a received speech+noise mixture, or of distorted speech, even with the relatively sophisticated enhancement algorithms that modern hearing aids are capable of running. A clever way around this problem might be for the sender to add extra information to the original speech signal, before noise or distortion is added. The receiver (e.g., a hearing aid) would use this to assist speech enhancement.
Funding: Marie Sklodowska-Curie fellowship
2. Broadcast Quality End-to-end Speech Synthesis
Advances in neural networks made jointly in the fields of automatic speech recognition and speech synthesis, amongst others, have led to a new understanding of their capabilities as generative models. Neural networks can now directly generate synthetic speech waveforms, without the limited quality of a vocoder. We have made separate advances, using neural networks to discover representations of spoken and written language that have applications in lightly-supervised text processing for almost any language, and for adaptation of speaker identity and style. The project will combine these techniques into a single end-to-end model for speech synthesis. This will require new techniques to learn from both text and speech data, which may have other applications, such as automatic speech recognition.
Funding: EPSRC Industrial CASE award (in collaboration with the BBC)
3. Automatic Extraction of Rich Metadata from Broadcast Speech (in collaboration with the BBC)
The research studentship will be concerned with automatically learning to extract rich metadata information from broadcast television recordings, using speech recognition and natural language processing techniques. We will build on recent advances in convolutional and recurrent neural networks, using architectures which learn representations jointly, considering both acoustic and textual data. The project will build on our current work in the rich transcription of broadcast speech using neural network based speech recognition systems, along with neural network approaches to machine reading and summarisation. In particular, we are interested in developing approaches to transcribing broadcast speech in a way appropriate to the particular context. This may include compression or distillation of the content (perhaps to fit in with the constraints of subtitling), transforming conversational speech into a form that is more easy to read as text, or transcribing broadcast speech in a way appropriate for a particular reading age.
Funding: EPSRC Industrial CASE award (in collaboration with the BBC)
(2017-04-20) Postdoc for project LaDyCa, Sorbonne, Paris
Applicants must have a PhD in linguistics as well as publications in their field of specialization. Independent research experience in one or several of the core areas of the LaDyCa project (i.e. language dynamics, linguistic typology, sociolinguistics, geolinguistics, dialectology & dialectometry) is expected. An experience in working with scholars of diverse backgrounds, e.g. linguists, sociologists, anthropologists, historians and, to some extent, mathematicians or statisticians would be greatly appreciated.The project will be funded by the IDEX (?Initiative d?Excellence?) consortium of Sorbonne Universités, France, in partnership with Ilia State University, Tbilisi, Georgia. Apart from an efficient and fluent command of English and/or French, for collegial relations. with an international team of scholars, applicants should have a good command of Georgian (written & oral skills); efficient reading skills in Russian would be an asset too. A good command of database software, and previous training or experience in computational linguistics would be also appreciated. A strong performing ability in entering data and in designing linguistic databases would be an asset. Applications should include a statement of interest (letter of motivation), giving accurate details on the applicant?s skills corresponding to the aim of the LaDyCa project, and how (s)he plans to process data with computing tools and gather information on the ecological, historical and social context of linguistic diversity in the Caucasus. (S)he will also provide a CV including a list of publications, a copy of the PhD certificate, and the names and e-mail addresses of two referees. Applications should be sent as a single PDF file to the e- mail addresses below, entitled ?Application_LaDyCa_PostDoc?:
The position is available from July 2017 to June 2018.The duration of employment is intended to last one year. Net salary: around 2100 euros per month.
L'équipe Expression de l'IRISA ouvre un poste de doctorant en informatique sur le sujet 'caractérisation de registres de langue par extraction de motifs séquentiels' dans le cadre du projet ANR TREMoLo.
Domaines : traitement automatique des langues et fouille de données.
Détails de l'offre : http://www.irisa.fr/fr/offres-theses/caracterisation-registres-langue-extraction-motifs-sequentiels
Date limite de candidature : vendredi 2 juin.
Dossier de candidature (* : éléments obligatoires) :
- CV détaillé*
- lettre de motivation*
- relevés de notes (avec classement si possible)*
- contacts pour recommandation*
- rapport(s) de stage recherche (si applicable).
Envoyer à : del.battistelli@gmail.com, nicolas.bechet@irisa.fr, gwenole.lecorve@irisa.fr.
(2017-05-05) Post-doctoral positions in Multimodal Behavior Analysis: Speech, Vision and Healthcare, CMU, Pittsburgh, PA, USA
Post-doctoral positions in Multimodal Behavior Analysis: Speech, Vision and Healthcare
Carnegie Mellon University, School of Computer Science
Multiple post-doctoral positions are available in the School of Computer Science at Carnegie Mellon University. We are seeking creative and energetic applicants for two-year postdoctoral positions. The positions include a competitive salary with full benefits and travel resources.
Candidates must have a strong research track record for one or more of the following topics: (1) speech and paralinguistic processing for affect, emotion and human behavior analysis, (2) automatic recognition of facial expressions, gestures and human visual activities, (3) multimodal machine learning algorithms for text, audio and video, (4) technologies to help clinicians with mental health diagnoses and treatments.
Required
PhD in computer science or related field (at the time of hire)
International applicants welcome! No US citizenship requirement.
Desired
Publications in top machine learning, speech processing and/or computer vision conferences and journals.
Research involving clinical patients with mental health disorders (e.g., depression, schizophrenia, suicidal ideation)
Experience mentoring graduate and undergraduate students
Job details
Preferred start date: September 1st, 2017 (negotiable)
Candidate will work under the supervision of Dr. Louis-Philippe Morency, CMU MultiComp Lab’s director
Competitive salary with full benefits and travel resources.
How to apply
Email applications should be sent to morency@cs.cmu.edu with the title “Postdoc application”, preferably before June 12th, 2017. The email should include:
a brief cover letter (with expected date of availability),
(2017-05-10) CDI Ingénieur docteur en informatique ou sciences du langage, LNE, Trappes, France
Ingénieur docteur en informatique ou sciences du langage
CDI – TRAPPES
Référence:AP/TAI/DE
L’entreprise: WWW.LNE.FR
Leader dans l’univers de la mesure et des références, jouissant d’une forte notoriété en France et à l’international, le LNE soutient l’innovation industrielle et se positionne comme un acteur important pour une économie plus compétitive et une société plus sûre. Au carrefour de la science et de l’industrie depuis sa création en 1901, le LNE offre son expertise à l’ensemble des acteurs économiques impliqués dans la qualité et la sécurité des produits.
Pilote de la métrologie française, notre recherche est au cœur de notre mission de service public et constitue un facteur fondamental au soutien de la compétitivité des entreprises.
Nous avons à cœur de répondre aux exigences des industriels et du monde académique, pour des mesures toujours plus justes, effectuées dans des conditions de plus en plus extrêmes ou sur des sujets innovants tels que les véhicules autonomes, les nanotechnologies ou la fabrication additive.
Le LNE en quelques chiffres: 700 collaborateurs.
5 métiers (la mesure, les essais, la certification, la formation et la R&D).
8 domaines d’intervention (Métrologie, Santé, Bâtiment, Environnement, Energie, Transports, Sécurité et Défense, Biens de consommation).
55 000 m2 de laboratoires (dont 10 000m2 à Paris et 45 000m2 à Trappes).
7 implantations (2 sites en Ile de France, 2 délégations régionales à Poitiers et Nîmes, 1 antenne à St Etienne, 2 filiales à Washington, Hong Kong).
9000 clients.
Missions :
Le docteur sera intégré à une équipe de 4 ingénieur-docteurs qui encadrent différents stagiaires et doctorants. Cette équipe est historiquement spécialiste de l’évaluation des systèmes de traitement de l’information multimédia (transcription de parole, reconnaissance du locuteur, dialogue, traduction…). Elle s’ouvre aujourd’hui à de nouveaux enjeux que sont l’évaluation des systèmes d’intelligence artificielle en général (robotique, smart-grid, domaine de la défense, véhicule autonome…).
Le docteur se verra attribuer les missions suivantes :
Le développement de la R&D en évaluation de systèmes de traitement de la parole et du langage
Définition de nouvelles métriques
Analyse de corpus
Publication de résultats scientifiques
Mise en place de protocoles perceptifs
Contribution au montage et déroulement de projets de recherche européens et nationaux.
Animation des campagnes d’évaluation
Aide aux équipes participantes pour l’utilisation des outils du LNE
Contrôle formel des données
Scoring des systèmes
Organisation de rencontres scientifiques et industrielles
Rédaction des rapports d’évaluation
Encadrement de stagiaires, post-docs
Profil :
Titulaire d’un doctorat en Informatique ou Sciences du langage, vous avez des compétences en traitement automatique de la langue ou en linguistique de corpus. Vous maitrisez également la programmation (R ou S, C++, PYTHON).
Vous êtes doté de bonnes qualités rédactionnelles et relationnelles. Vous avez une bonne communication orale et vous aimez travailler en collaboration avec votre équipe et les clients.
Vous avez un anglais vous permettant une communication professionnelle.
Déplacements en région parisienne, 1 jour par semaine et dans le monde 1 fois par an.
Pour déposer votre candidature :envoyer CV+LM à recrut@lne.fr – réf AP/TAI/DE
(2017-05-10) Open Rank Research Scientist, Spoken and Multimodal Dialog Systems, ETS, San Francisco, CA, USA
Open Rank Research Scientist, Spoken and Multimodal Dialog Systems
ETS (Educational Testing Service) is a global not for profit organization whose mission is to advance quality and equity in education. With more than 3,400 global employees, we develop, administer and score more than 50 million tests annually in more than 180 countries.
Our San Francisco Research and Development division is seeking a Research Scientist for our Dialog, Multimodal, and Speech (DIAMONDS) research center. The center’s main focus is on foundational research as well as on development of new capabilities to automatically score spoken, interactive, and multimodal test responses in conversational settings in a wide range of ETS test programs, promote learning and other educational areas. This is an excellent opportunity to be part of a world-renowned research and development team and have a significant impact on existing and next generation spoken and multimodal dialog systems and their application to assessment and other areas in education.
Primary responsibilities include:
Developing and collaborating on interdisciplinary projects that aim to transfer techniques to a new context or scientific field. Successful candidates are self-motivated and self-driven, and have a strong interest in emerging conversational technology that can contribute to education in assessment and instructional settings.
Providing scientific and technical skills to conceptualize, design, obtain support for, conduct, and manage new research projects, grants, or parts of existing projects.
Generating or contributing to new or modified methods that support research on and development of spoken and multimodal dialog systems and related technologies relevant in assessment and instructional settings.
Designing and conducting scientific studies and functioning as an expert in the major facets of the projects: responding as a subject matter expert in presenting the results of acquired knowledge and experience.
Developing or assisting in developing proposals for external and internal research grants and obtain financial support for new or continuing research activities. Prepare initial and final proposal and project budgets.
Participating in dissemination activities through the publications of research papers in peer-reviewed journals and in the ETS Research Report series, the issuing of progress and technical reports, the presentation of seminars at major conferences and at ETS, or the use of other appropriate communication vehicles, including patents, books and chapters, that impact practice in the field or at ETS.
Depending on experience this position is open to entry level candidates as well as mid-level and senior level professionals.
REQUIREMENTS FOR A JUNIOR LEVEL POSITION
A Doctorate in computer science, linguistics, cognitive psychology or a related field is required. One year of research experience is required, in education is desirable. Experience can be gained through doctoral studies. Candidates should be very skilled in programming and be able to work effectively as a research team member.
REQUIREMENTS FOR A MID-LEVEL POSITION
A Doctorate in computer science, linguistics, cognitive psychology, or a related field is required. Research experience in education is desirable. Candidates should be very skilled in programming and be able to work effectively as a research team member. Three years of progressively independent substantive research in the area of computer science, linguistics, cognitive psychology, or education are required.
REQUIREMENTS FOR A SENIOR-LEVEL POSITION
A Doctorate in computer science, linguistics, cognitive psychology, or a related field is required. Research experience in education is desirable. Candidates should be very skilled in programming and be able to work effectively as a research team member. Eight years of progressively independent substantive research in the area of computer science, linguistics, cognitive psychology, or education are required.
(2017-05-10) Research Scientist, Disney Research, Pittsburgh, PA, USA
Position: Research Scientist
Focus Area: Autonomous Agents for Multimodal Character Interaction
Disney Research
Disney Research Pittsburgh is seeking applicants for a Research Scientist position, at either the junior or senior level, in Autonomous Agents. The research emphasis is on architecture to support the integration of natural language with character-based reasoning and behavior.
As part of The Walt Disney Company, Disney Research builds upon a rich legacy of innovation and technology leadership in the entertainment industry that continues to this day. Disney Research was launched in 2008 offering the best attributes of academia and industry with the goal of driving value across the company through technological innovation. Our research covers a broad range of exciting and challenging applications that are experienced daily by millions of people around the world.
Our staff interacts directly with all core business areas of The Walt Disney Company including Theme Parks and Imagineering, Consumer Products, our Live Action and Animation Studios, and Media Networks. We publish our research and are actively engaged with the global research community. Our researchers collaborate closely with co-located academic institutions.
We are seeking applicants in the following areas:
· Agent architectures for language-based character interaction.
· AI and machine learning methods for autonomous, semantically-rich character behavior
Duties:
· Drive value for Disney through groundbreaking research and innovation
· Lead a research group with post-doctoral researchers, interns, and external collaborators
· Publish results and patent inventions in multimodal interaction
· Participate in conferences, workshops and academic-industrial events
· Develop a strong network of business partners within the company
Required Qualifications:
· Ph.D. in Computer Science or equivalent
· Proven track record of developing autonomous, integrated agents with real-time NL components.
· Experience with both symbolic and statistical machine learning methods as applied to modeling semantics, action, or behavior
· Possess strong technical presentation skills and able to clearly communicate with technical and non-technical audiences
Desired Qualifications:
· Experience in interaction design for entertainment
· Background in NLP (e.g., relationship extraction, word sense disambiguation, narrative generation) desirable
To apply:
Please email careers@disneyresearch.com. Please use DRP-RS-NLP-2017 in your subject line. If you're interested in the position or for any further information, please contact Jill Lehman (jill.lehman@disneyresearch.com).
Speechmatics is a leader in automatic speech recognition (ASR). Using proprietary technology, we have built one of the most accurate ASR systems in the world, with a vision to power a voice-enabled economy. We are already working at a time when the global economy is actively adopting all types of speech-related technologies. In developing our technology we combine our years of experience, the latest developments in the field and our own focus on cutting-edge research to produce a world-class service.
In the office, we pride ourselves on a relaxed but productive environment whilst we stay in touch with the progress of others by attending both academic and commercial conferences and have fun together with regular outings (in the past we have been punting, go-karting, attended a cooking workshop and played bubble football...).
We are expanding rapidly and are seeking more people in the coming months to help us keep pushing the boundaries of speech recognition. This is an opportunity to join a high growth team and form a major part of its future direction.
The Opportunity
We are looking for a talented speech scientist to help us build the best speech technology for anybody, anywhere, in any language. You will be a part of a team that is working on our core ASR capabilities to improve our speed and accuracy and develop novel features so that we can support all languages. Your work will feed into ‘Auto-Auto’, our ground-breaking framework to support the building of ASR models, and hence the delivery of every language pack published by the company. You will be responsible for keeping our system the most accurate and useful commercial speech recognition available.
Because you will be joining a small team, you will need to be a team player who thrives in a fast paced environment, with a focus on rapidly moving research developments into products. Bringing skills to the team is as important as a can-do attitude. We strongly encourage versatility and knowledge transfer within the team, so we can share efficiently what needs to be done to meet our commitments to the rest of the company.
Key Responsibilities
Ensuring that our speech recognition meets or exceeds that published by others
Leading the extension of our ML framework so that we can build any language
Experience
Essential
MSc, PhD or equivalent experience in the academic aspects of speech recognition
Several years practical experience in speech recognition, covering all aspects (acoustic, pronunciation and language modelling as well as decoders/search)
Experience working with standard speech and ML toolkits, e.g. Kaldi, KenLM, TensorFlow, etc.
Solid programming skills with Python and / or C/C++
Experience using Unix/Linux for big data
Desirable
PhD degree
Experience of team leadership and line management
Experience of working in an Agile framework
Expertise in modern speech recognition, including WFSTs, lattice processing, neural net (RNN / DNN / LSTM), acoustic and language models, Viterbi decoding
Comprehensive knowledge of machine learning and statistical modelling
Experience in deep machine learning and related toolkits, e.g. Theano, Torch, etc.
Deep expertise in Python and/or C++ software development
Experience working effectively with software engineering teams or as a Software Engineer
Salary
We offer a competitive salary, bonus scheme, pension contribution matching (up to 5%) and a generous EMI share option scheme. We also have several additional benefits including holiday purchase, massages, fully stocked beer fridge, Cyclescheme, fruit boxes and many more.
The overall package will depend on your motivations and level of experience.
Speechmatics is a leader in automatic speech recognition (ASR). Using proprietary technology, we have built one of the most accurate ASR systems in the world, with a vision to power a voice-enabled economy. We are already working in the world at a time when the global economy is actively adopting all types of speech-related technologies. In developing our technology we combine our years of experience, the latest developments in the field and our own focus on cutting-edge research to produce a world-class service.
In the office, we pride ourselves on a relaxed but productive environment whilst we stay in touch with the field by attending both academic and commercial conferences and have fun together with regular team events (in the past we have been punting, go-karting, attended a cooking workshop and played bubble football...).
We are expanding rapidly and are seeking more people in the coming months to help us keep pushing the boundaries of speech recognition. This is an opportunity to join a high growth team and form a major part of its future direction.
The Opportunity
You will be joining the ‘Languages’ team within Speechmatics, focussing on two key goals. We maintain and develop Auto-Auto, our ground-breaking framework to support the building of languages for use in ASR. And we use it to build new language models.
We are looking for an experienced Software Development Engineer to join us. As a member of the team, you will be working on the development, maintenance and expansion of our pipeline, and participating in building and solving the challenges of a growing language portfolio. You will have significant influence on implementing or integrating new features, drive the system architecture, and spearhead the best practices that enable a quality product.
Auto-Auto is core to our business and by working on it you will have a chance to build something that will be used in businesses and homes worldwide. Working in a rapidly growing start-up also means opportunities to contribute to other projects, depending on the candidate’s background and skills.
If you are a talented, detail-oriented engineer with a solid software development foundation and a commitment to deliver the best possible technology solutions, then we want to hear from you!
Key Responsibilities
Delivering high quality, maintainable and robust code on time, as part of a team.
Executing projects and developing against an outlined design.
Developing pragmatic solutions and building flexible systems without over-engineering.
Involvement at all stages of the software development cycle, including designing and developing new architectural systems and improvements, and QA processes.
Participation in estimation and sprint planning in an agile environment.
Participation in delivering new language models for the ASR engine.
Working closely with other technical teams and product team to deliver on the company’s technical vision.
Experience
Essential
Bachelor's Degree in Computer Science or related field.
Professional experience in software development.
Computer Science fundamentals in object-oriented design, data structures, algorithm design, problem solving, and complexity analysis.
Knowledge of professional software engineering practices & best practices for the full software development life cycle, including coding standards, code reviews, source control management, build processes, testing, and operations.
Excellent Python skills.
Good Linux skills.
Experience of working within a team to deliver and run high quality systems.
Desirable
Master's degree in Computer Science or related field.
Demonstrable professional experience in software development.
Proficiency in C and C++ (ideally with strong STL and Boost experience).
Strong skills and experience in cloud-based software development, preferably AWS:
Working with distributed and/or clustered systems.
Building and running horizontally scaling architectures.
Using cloud-based queueing, messaging, monitoring and storage techniques.
Experience in flow-based programming.
Familiarity with statistical models and data mining algorithms.
Analytical with a data-driven approach to making decisions and attention to detail.
Previous experience with Natural Language Processing techniques.
Comfortable collaborating with teams with very different technical skills, and non-technical teams.
Salary
We offer a competitive salary, bonus scheme, pension contribution matching (up to 5%) and a generous EMI share option scheme. We also have several additional benefits including holiday purchase, massages, fully stocked beer fridge, Cyclescheme, fruit boxes and many more.
(2017-05-29) PhD & Post-Doc Research positions in Speech Signal Processing and Electronic Design, Autonomous University of Zacatecas, Zacatecas, Mexico
PhD & Post-Doc Research positions in Speech Signal Processing and Electronic Design
Place: Autonomous University of Zacatecas, Zacatecas, Mexico
Duration: PhD (3 years) / Post-Doc (1 year)
Start: PhD / Post-Doc (January 10th, 2018)
Benefits: - Economical support according to experience - Health insurance from the Mexican Social Security Institute - Round-trip international airfare at the beginning and end.
Position description: Department of Signal Processing and Acoustics, Autonomous University of Zacatecas, is looking for candidates for: fully-funded PhD and Post-doc positions in Signal Processing, Filtering Design, Embedded Systems, Speech Recognition and Synthesis. The signal processing group (led by Dr. Hamurabi GamboaRosales) at Autonomous University of Zacatecas works on algorithm designing, signal processing, electronic design, machine learning, probabilistic modeling in speech recognition and synthesis. The group belongs also to the National Laboratory in embedded systems, advanced electronic design and Microsystems. We are looking for outstanding candidates to join our research group as PhD students and Post-doc researchers to work on any of our research themes, for example: • Digital Signal processing • Optimal Filtering • FPGA’s • Microsystems • large-vocabulary speech recognition and text- to-speech synthesis • ASR in noisy environments
Candidate Profile for PhD / Post-Doc: The candidate will have: Master’s / PhD degree, as required by the program for which is requested, in digital signal processing, electronic design, speech signal processing, acoustics, machine learning, computer science, electrical engineering, psychology or a related discipline. Background in signal processing or electronic design (FPGAs). Good programming skills in Java, C/C++, Python or Matlab.
Contact: Interested applicants can contact PhD Hamurabi Gamboa-Rosales for more information or directly email a candidacy letter including Curriculum Vitae, a list of publications and a statement of research interests. Email: hamurabigr@uaz.edu.mx ; hamurabigr@hotmail.com Telephone MX: +52 (1) 492-121-6787
(2017-06-01) Appel à chercheurs 2017-2018 à l'INA Paris France
Appel à chercheurs 2017-2018
Nouveaux dispositifs de soutien à la recherche à l’Ina :
Chercheurs associés et bourses de recherche
Afin d’encourager le développement de travaux scientifiques menés à partir de ses fonds et des outils d’analyse qu’il développe, l'Ina a décidé de créer en 2017 deux nouveaux dispositifs de soutien à la recherche et à la valorisation scientifique de ses collections :
l’octroi d’un statut de chercheur associé à l’Ina
l’attribution de bourses de recherche
Par ces dispositifs, l’Ina entend accompagner des doctorants et des chercheurs dans la réalisation de projets de recherche originaux et innovants portant sur (ou faisant appel à) ses collections, ou portant sur l’analyse ou le traitement des images et/ou des sons et/ou de données associées.
L’Institut offre aux chercheurs sélectionnés un accueil privilégié, assorti de divers soutiens matériels.
Ces nouveaux dispositifs sont complémentaires des prix de l’Inathèque créés en 1997, et ajoutent un nouveau volet à la politique scientifique de l’Institut.
(2017-06-023) Vacataires à la police technique et scientifique, Ecully (Lyon), France
Le service audio de la Police Technique et Scientifique (Ecully, près de Lyon, France) recherche des vacataires pour effectuer un travail de segmentation et de correction d'alignement automatique dans le cadre d'études phonétiques. Le profil suivant est recherché:
- un intérêt pour la linguistique ou pour les langues
- une bonne maitrise de l'informatique et des nouvelles technologiques
- une connaissance du logiciel Praat sera appréciée
Les vacations peuvent commencer dès que possible et peuvent se poursuivre jusqu'en octobre.
Pour plus d'informations, merci d'envoyer un mail à l'adresse suivante ptsvox@gmail.com, avec vos coordonnées.
(2017-06-06) Post-doctoral Research Associate in Advanced Deep Neural network Architectures for ASR , Univ. of Crete, Greece
Department of Computer Science, University of Crete, Greece Post-doctoral Research Associate in Advanced Deep Neural network Architectures for ASR (Fixed Term) SALARY: €24000-€28000 per year CLOSING DATE: 30 June 2017 REFERENCE: ASR1 TO APPLY: Send detailed CV, a motivation letter and 3 major publications to yannis@csd.uoc.gr In the past few years, Deep Neural Networks (DNNs) have achieved tremendous success for many supervised machine learning tasks, including acoustic modelling for Automatic Speech Recognition (ASR). Advanced models such as Convolutional Neural Networks (CNNs) and Long Short Term Recurrent Neural Networks (LSTMs) have contributed to recent empirical breakthroughs. Network depth has played perhaps the most important role in these successes. However, increased depth represents challenges in the optimization of the network and despite the efforts to overcome these challenges some of the optimization issues are still important resistant. Advanced networks such as highway networks and (wide) residual networks seems to offer solutions to these issues. This position represents an ideal opportunity to work in or move into advanced deep neural networks, as it will involve collaborating widely across academia and industry, and working on one of the most pressing research areas of machine learning for the development of robust ASR systems. Based in Heraklion Crete the post will be with Prof. Yannis Stylianou and Dr. Vassilis Tsiaras as part of the speech processing group within the Department of Computer Science at the University of Crete. You will explore a rich set of network architectures and thoroughly examine how several different aspects affect the accuracy of ASR. The work will be performed within the framework of advanced deep neural network architectures for various signal processing tasks including 1D and 2D signals. The focus of the post will be to perform various experiments with well-known architectures, explore and suggest modifications, process and reshape knowledge from various signal processing/classification tasks towards speech processing for the purpose of ASR. Outcomes will directly feed into improvements of ASR systems in-house working with state-of-the art ASR tasks (i.e., CHiME4, REVERB, etc) and of our industrial partners using real-life data. The post involves travel to international conferences and project meetings with our academic and industrial partners. There will be the possibility to co-advise doctoral students and potentially other teaching opportunities. Applicants should have a doctorate in speech signal processing area for ASR, computer science, applied mathematics or related field and ideally a strong background in deep learning and mathematics. Knowledge of deep learning systems such as Tensorflow or Theano etc and ASR systems like Kaldi are an advantage. Proficiency in computer programming in C and/or Python are expected. Informal inquiries should be directed to Prof. Yannis Stylianou by email, yannis@csd.uoc.gr Fixed term: In the first instance, the funding supporting the post is for two years. We are expecting project extension which will provide funding for a further 7-12 months for this post. Interviews are expected to take place the week commencing 10th July 2017. Expected start date: September 2017, however earlier and later start dates will be considered. To apply, please send detailed CV, a motivation letter and 3 major publications of yours to: yannis@csd.uoc.gr (Prof. Yannis Stylianou)
(2017-06-06) Post-doctoral Research Associate in Data Augmentation in the context of Deep Neural network ASR, Univ.of Crete, Greece
Department of Computer Science, University of Crete, Greece
Post-doctoral Research Associate in Data Augmentation in the context of Deep Neural network ASR
(Fixed Term)
SALARY: €24000-€28000 per year
CLOSING DATE: 30 June 2017
REFERENCE: ASR2
TO APPLY: Send detailed CV, a motivation letter and 3 major publications to yannis@csd.uoc.gr
In the past few years, Deep Neural Networks (DNNs) have achieved tremendous success for many supervised machine learning tasks, including acoustic modelling for Automatic Speech Recognition (ASR). Advanced models such as Convolutional Neural Networks (CNNs) and Long Short Term Recurrent Neural Networks (LSTMs) have contributed to recent empirical breakthroughs. However, deep learning methods are quite demanding in the amount of data for training an acoustic model for ASR and as a result significant amounts of transcribed data has become available for training use. But data transcription is a quite expensive and time consuming process. On the other hand, just adding data recorded in real-world conditions puts serious constraints on the efficient training of the acoustic models. Various works on data augmentation show that word error rate (WER) can be significantly reduced if proper augmented data are processed.
This position represents an ideal opportunity to work in or move into data augmentation research area in the context of advanced deep neural networks for ASR, as it will involve collaborating widely across academia and industry, and working on one of the most pressing research areas of machine learning for the development of robust ASR systems.
Based in Heraklion Crete the post will be with Prof. Yannis Stylianou and Dr. George Kafentzis as part of the speech processing group within the Department of Computer Science at the University of Crete. You will design and develop smart approaches for spoken data augmentation for the purpose of multi-condition training of deep learning-based ASR systems. The work will be performed within the framework of advanced deep neural network architectures for various ASR tasks. The focus of the post will be to perform various experiments with spoken data generation, explore and suggest modifications, process and reshape knowledge from various signal processing for the purpose of ASR. Outcomes will directly feed into improvements of ASR systems in-house working with state-of-the art ASR tasks (i.e., AURORA-4, CHiME4, REVERB, etc) and of our industrial partners using real-life data.
The post involves travel to international conferences and project meetings with our academic and industrial partners. There will be the possibility to co-advise doctoral students and potentially other teaching opportunities.
Applicants should have a doctorate in speech signal processing area for ASR, statistical speech synthesis and voice conversion, audio signal processing, computer science, applied mathematics or related field and ideally a strong background in deep learning and mathematics. Knowledge of deep
learning systems such as Tensorflow or Theano etc and ASR systems like Kaldi are an advantage. Proficiency in computer programming in C and/or Python are expected.
Informal inquiries should be directed to Prof. Yannis Stylianou by email, yannis@csd.uoc.gr
Fixed term: In the first instance, the funding supporting the post is for two years. We are expecting project extension which will provide funding for a further 7-12 months for this post.
Interviews are expected to take place the week commencing 10th July 2017.
Expected start date: September 2017, however earlier and later start dates will be considered.
To apply, please send detailed CV, a motivation letter and 3 major publications of yours to: yannis@csd.uoc.gr (Prof. Yannis Stylianou)
(2017-06-18) Proposition de thèse CIFRE en Informatique, traitement automatique du langage naturel
Proposition de thèse CIFRE en Informatique, traitement automatique du langage naturel
Société Calystene et Laboratoire d'Informatique de Grenoble
Calystene propose une thèse CIFRE dans le cadre d'une collaboration avec l'équipe GETALP du laboratoire d'Informatique de Grenoble.
Début de la thèse : Dès que possible. Localisation : le poste est basé à Grenoble - Eybens Le salaire est de 30 000 keuros brut annuel environ.
Description du sujet :
Avec l'évolution des systèmes d'information des établissements de santé, les praticiens sont incités à saisir de plus en plus d'informations de manière numérique à travers différents logiciels spécialisés. La prise en main de ces logiciels implique toujours une étape de formation et d?adaptation qui est problématique pour certains clients de Calystene. Ainsi, Calystene souhaite rendre naturelle la saisie de prescription médicamenteuse sur son logiciel FUTURA SMART DESIGN® pour les praticiens, notamment ceux extérieurs aux établissements de santé. En effet, ces intervenants, précieux mais ponctuels, n'ont pas le temps de se former aux interfaces des logiciels et n'ont pas d'accès immédiat à un poste de travail (cas notamment des médecins de ville qui interviennent en maison de retraite). Des prescriptions sont donc régulièrement rédigées sur papier et n'entrent pas dans le système d'information, ce qui est dommageable pour la gestion des soins.
Pour répondre à ce problème, ce doctorat vise à définir des modèles d'interaction en langage naturel et leur mise en ?uvre dans la plateforme FUTURA SMART DESIGN® de Calystene, afin de proposer une saisie naturelle d'informations par écrit et/ou orale de prescriptions médicales sur un terminal mobile et pour la résolution intuitive et intelligente d'erreurs ou d'information manquante. Le prototype, baptisé « Intelligent Prescription Completion » permettra de se rapprocher le plus possible du langage de prescription des médecins. Les lignes d?ordonnance seront saisies soit par dictée vocale, reconnaissance d?écriture sur tablette ou saisie classique via le clavier. L?ordonnance sera saisie en langage quasi-naturel (le langage des prescriptions médicales) et la structuration et validation des données (nécessaire aux contrôles et à la gestion) sera réalisée en temps réel par un algorithme développé par Calystene.
Pour atteindre ces objectifs, la solution s?appuiera sur une approche de type dialogue [1] où la sémantique de l'information est d'abord extraite de l?énoncé [2] --- qu'il soit écrit ou acquis à partir de la parole [3] --- puis analysée par la logique métier afin de valider, suggérer des modifications ou provoquer des demandes de complément d?information à la manière d'un chatbot. La thématique du doctorat se situera donc dans le domaine de la compréhension automatique du langage naturel et de la parole et du raisonnement automatique avec un fort aspect prototypage et validation. Le/la candidat(e) sera encouragé(e) à publier ses progrès dans les grandes conférences du domaine en TALN (ACL,Interspeech) et en IA appliquée à la médecine (AIME, AMIA).
[1] S. Young, M. Ga?i?, S. Keizer, F. Mairesse, J. Schatzmann, B. Thomson, and K. Yu, ?The hidden information state model: A practical framework for pomdp-based spoken dialogue management,? Computer Speech & Language, vol. 24, no. 2, pp. 150?174, 2010 [2] Xu H, Stenner SP, Doan S, Johnson KB, Waitman LR, Denny JC. MedEx: a medication information extraction system for clinical narratives. Journal of the American Medical Informatics Association?: JAMIA. 2010;17(1):19-24. [3] M. Vacher, S. Caffiau, F. Portet, B. Meillon, C. Roux, E. Elias, B. Lecouteux, P. Chahuara. Evaluation of a context-aware voice interface for Ambient Assisted Living: qualitative user study vs. quantitative system evaluation ACM - Transactions on Speech and Language Processing, Association for Computing Machinery, 2015, Special Issue on Speech and Language Processing for AT (Part 3), 7 (issue 2), pp.5:1-5:36.
Profil du candidat recherché : Possédant ou terminant un Master 2 Recherche en informatique ou en TALN, vous souhaitez préparer un doctorat CIFRE en entreprise en liaison avec un laboratoire de recherche. Vous êtes passionné par les technologies de l?information et de la communication, vous avez de bonnes connaissances en développement Android/iPhone. Vous avez une formation et une expérience dans l'étude et/ou le développement de traitement automatique du langage naturel. Des connaissances en apprentissage automatique et en acquisition de corpus seraient un plus.
Merci d'envoyer un CV + une lettre de motivation + lettres de recommandation à Jean-Marc BABOUCHKINE (jm.babouchkine@calystene.com) et François Portet (francois.portet@imag.fr)
(2017-06-19) PhD position in Computational Linguistics for Ambient Intelligence, Grenoble, France
Keywords: Natural language understanding, decision support system, smart home
The Laboratoire d'Informatique de Grenoble (LIG) of the University Grenoble Alpes, Grenoble, France invites applications for a PhD position in Computational Linguistics for Ambient Intelligence.
University of Grenoble Alpes is situated in a high-tech city located at the heart of the Alps, in outstanding scientific and natural surroundings. It is 3h by train from Paris ; 2h from Geneva and is less than 1h from Lyon international airport.
The position starts in September 2017 and ends in July 2020 and is proposed in the context of the national project Vocadom (http://vocadom.imag.fr/) whose aim is to build technologies that make natural hand-free speech interaction with a home automation system possible from anywhere in the home even in adverse conditions [Vacher2015].
The aim of the PhD will be to build a new generation of situated spoken human machine interaction where uttered sentences by a human are understood within the context of the interaction in the home. The targeted application is a distant speech hand free and ubiquitous voice user interface to make the home automation system react to voice commands [Chahuara2017]. The system should be able to process possibly erroneous outputs from an ASR system (Automatic Speech Recognition) to extract meaning related to a voice command and to decide about which command to execute or to send a relevant feed-back to the user. The challenge will be to constantly adapt the system to new lexical phrases (no a priori grammar), new situations (e.g., unseen user, context) and change in the house (e.g., new device, device out of order). In this work, we propose to extend classical S/NLU (Natural Language Understanding) approaches by including non-linguistic contextual information in the NLU process to tackle the ambiguity and borrow zero-shot learning techniques [Ferreira2015] to extend the lexical space on-line. Reinforcement learning is targeted to adapt the models to the user(s) all along the use of the system [Mnih2015]. The candidate will be strongly encouraged to publish their progress to the main events of the field (ACL, Interspeech, Ubicomp). The PhD candidate will also be involved in experiments including real smart-home and real users (elderly people and people with visual impairment) [Vacher2015].
REFERENCES :
[Mnih2015] Mnih, Kavukcuoglu et al. Human-level control through deep reinforcement learning. Nature 518,529?533
[Chahuara2017] Chahuara, F. Portet, M. Vacher Context-aware decision making under uncertainty for voice-based control of smart home Expert Systems with Applications, Elsevier, 2017, 75, pp.63-79. [Ferreira2015] E Ferreira, B Jabaian, F Lefevre Online adaptative zero-shot learning spoken language understanding using word-embedding Acoustics, Speech and Signal Processing (ICASSP), 2015
[Vacher2015] M. Vacher, S. Caffiau, F. Portet, B. Meillon, C. Roux, E. Elias, B. Lecouteux, P. Chahuara. Evaluation of a context-aware voice interface for Ambient Assisted Living: qualitative user study vs. quantitative system evaluation. ACM - Transactions on Speech and Language Processing, Association for Computing Machinery, 2015, pp.5:1-5:36.
JOB REQUIREMENTS AND QUALIFICATIONS
- Master?s degree in Computational Linguistics or Artificial Intelligence (Computer Science can also be considered) - Solid programming skills, - Good background in machine learning, - Excellent English communication and writing skills, - Good command of French (mandatory), - Experience in experimentation involving human participants would be a plus - Experience in dialogue systems would be a plus plus
Applications should include:
- Cover letter outlining interest in the position - Names of two referees - Curriculum Vitae (CV) (with publications if applicable) - Copy of the university marks (grade list)
(2017-06-20) Full-time post-Doctoral researcher position at LORIA Nancy, France
Loria a computer science lab in Nancy - France, has 12 months funded full-time post-Doctoral researcher position starting on October 2017. The post-doctoral position is funded by AMIS (Access Multilingual Information OpinionS), a Chist-Era project (http://deustotechlife.deusto.es/amis/).
The topic of the post-doc is the automatic comparison of multilingual opinions in videos. Two videos in two different languages concerning the same topic have to be compared. One of the videos is summarized and translated to the language of the second one. This last one is summarized and then the opinions of the two original videos are compared in terms of emotional labels such as: anger, disgust, fear, joy, sadness, surprise... They should be compared also in terms of basic sentiments.
Social network will be used in order to reinforce the analysis of the contents in terms of opinions and sentiments.
AMIS group will make available the summary of videos in terms of text. The candidate will work on NLP, but skills in video analysis will be appreciated.
The applicant will contribute also to other tasks in collaboration with other partners of AMIS project.
The successful candidate will join the SMarT research team, he will be supervised by Prof. Kamel Smaïli, Dr D. Langlois and Dr D. Jouvet. The applicant will work also with Dr O. Mella and Dr D. Fohr.
Location: Loria - Nancy (France) Duration: October 2017 ? September 2018 Net salary: from 1800 Euros to 2400 Euros per month.
The ideal applicant should have:
* A PhD in NLP, opinion and sentiment mining or other strongly related discipline. * A very solid background in statistical machine learning. * Strong publications. * Solid programming skills to conduct experiments. * Excellent level in English.
Applicants should send to smaili@loria.fr:
* A CV * A cover letter outlining the motivation * Three most representative papers
(2017-06-20) Thèse CIFRE at Orange Labs, Lannion, France
Orange Labs propose une thèse CIFRE en Informatique en lien avec les domaines suivants : Apprentissage automatique, Prédiction structurée et Traitement automatique du langage naturel.
Cette thèse se place dans le cadre d'une collaboration avec l?équipe Expression de l?IRISA à Lannion.
La description du sujet est disponible à cet emplacement :
Profil des candidats : Les candidat(e)s doivent être titulaire d'un Master recherche en informatique, statistique, traitement du signal. Les formations mixtes mathématique/informatique sont privilégiées.
(2017-06-26) Language Resources Project Manager -Junior position at ELDA Paris France
The European Language resources Distribution Agency (ELDA), a company specialized in Human Language Technologies within an international context is currently seeking to fill an immediate vacancy for a Language Resources Project Manager ? Junior position. This yields excellent opportunities for young, creative, and motivated candidates wishing to participate actively to the Language Engineering field.
Language Resources Project Manager - Junior (m/f)
Under the supervision of the Language Resources Sales Manager, the Language Resources Project Manager ? Junior will be in charge of the identification of Language Resources (LRs) and the negotiation of rights in relation with their distribution.
The position includes, but is not limited to, the responsibility of the following tasks:
Identification of LRs and Cataloguing.
Negotiation of distribution rights, including interaction with LR providers, drafting of distribution agreements, definition of prices of language resources to be integrated in the ELRA catalogue.
LR Packaging and Archiving.
Designing and evaluating workflows for IPR clearance in the digital environment.
Profile:
Master?s degree or equivalent in Law and Computer science, with an awareness of Intellectual Property Rights and Data Protection issues in the digital environment.
The ideal candidate will have experience in computational linguistics, information science, knowledge management or similar fields.
Experience in project management and participation in European projects, as well as practice in contract and partnership negotiation at an international level, would be a plus.
Dynamic and communicative, flexible to combine and work on different tasks.
Ability to work independently and as part of a team.
Proficiency in English, with strong writing and documentation skills. Communication skills required in a French-speaking working environment.
Citizenship of (or residency papers) a European Union country.
All positions are based in Paris. Applications will be considered until the position is filled.
Salary is commensurate with qualifications and experience. Applicants should email a cover letter addressing the points listed above together with a curriculum vitae to:
ELDA 9, rue des Cordelières 75013 Paris FRANCE Fax : 01 43 13 33 30 Mail: job@elda.org
ELDA is acting as the distribution agency of the European Language Resources Association (ELRA). ELRA was established in February 1995, with the support of the European Commission, to promote the development and exploitation of Language Resources (LRs). Language Resources include all data necessary for language engineering, such as monolingual and multilingual lexica, text corpora, speech databases and terminology. The role of this non-profit membership Association is to promote the production of LRs, to collect and to validate them and, foremost, make them available to users. The association also gathers information on market needs and trends.
