![]() |
ISCApad #155 |
Saturday, May 14, 2011 by Chris Wellekens |
6-1 | (2010-11-11) Joint PhD Studentship in Speech-based Security over the Telephone (Avignon-Swansea) Fully-funded Joint PhD Studentship in Speech-based Security over the Telephone This Studentship is for a Joint Phd with the Speech & Image Research Group at Swansea University and Laboratoire Informatique d'Avignon at Université d’Avignon et des Pays de Vaucluse. This studentship is fully funded and worth up to £15,000 p.a. plus fees. Closing Date: Students are advised to apply as soon as possible. Who is eligible:
Further information: This studentship is directly linked to a new EU Eureka/FP7 programme the overall aim of which is to enhance security of transactions over the telephone using speech-based signal processing and new biometric approaches. Over the last 20 years both LIA and Swansea University have been at the forefront of international research in the area of speaker recognition How to Apply: Send your CV to: Dr John Mason j.s.d.mason@swansea.ac.uk or Dr. Jean-Francois BONASTRE jean-francois.bonastre@univ-avignon.fr
| |||||||||||||||||||||||||||||
6-2 | (2010-11-28) Post Doc at KUL Leuven Belgium Ph D Position or Post-Doc Position in Speech Recognition at K.U.Leuven
| |||||||||||||||||||||||||||||
6-3 | (2010-12-01) Opened positions/internships at Microsoft: French Linguists (M/F)
Opened positions/internships at Microsoft: French Linguists (M/F) MLDC – Microsoft Language Development Center, a branch of the Microsoft Product Group that develops Speech Recognition and Synthesis Technologies, situated in Porto Salvo, Portugal ( Be native or near native Have a university degree in Linguistics (with good computational skills), Computational Linguistics (Master’s or PhD) or related area Have an advanced level of English (oral and written) Have some experience in working with Speech Technology/Natural Language Processing/Linguistics, either in academia or in industry Have some computational ability – being able to run tools, being comfortable to work with Microsoft Office tools and having some programming fundamentals, though no programming is required Have team work skills Willing to work in Porto Salvo (near Lisbon) for the duration of the contract Willing to work in a multicultural and multinational team across the globe To apply, please submit your resume and a brief statement describing your experience and abilities to We will only consider electronic submissions. Deadline for submissions
| |||||||||||||||||||||||||||||
6-4 | (2010-12-06) These CIFRE at EDF France (in french) Financement : thèse CIFRE en traitement automatique de la langue parlée Sujet : L'extraction d'information à partir de données acoustiques de parole lance aujourd'hui un défi important aux chercheurs du traitement automatique de la parole: aller au-delà de la transformation du signal sonore en une simple suite de mots (transcription automatique) en y intégrant des informations sémantiques. Ces informations sémantiques sont notamment utilisées pour indexer les données acoustiques en fonction des thèmes abordés, des sujets traités, voire même des opinions exprimées sur ces sujets. De nombreuses recherches portées par la communauté TAL (traitement automatique des langues) abordent également la problématique de l'extraction d'information à partir de données textuelles de plus en plus variées et de sources hétérogènes (blogs, forum de discussions, etc.). En particulier, l'utilisation de techniques issues de l'indexation de documents textuels, appliquées à la transcription automatique de parole constitue un thème de recherche en plein essor qui soulève de nombreux problèmes scientifiques. Ces problèmes sont liés au traitement de la variabilité dans la parole, aux erreurs (ou au bruit) de la transcription automatique ainsi qu'aux modes d'expressions spontanés différents de ceux de l'expression écrite. Le sujet de thèse porte sur l'extraction d'information dans des données conversationnelles EDF à partir des transcriptions automatiques. Ce travail est à la fois concerné par l'extraction d'information à partir du texte et par l'analyse et la gestion des aspects liés à l'expression orale via le traitement des erreurs de transcription automatique. Les données conversationnelles EDF (centres d'appel, enquêtes de satisfaction) représentent un corpus riche fournissant un cadre de travail particulièrement motivant en termes de recherche fondamentale et d'enjeux applicatifs. La palette thématique et expressive est à la fois large et circonscrite par le cadre applicatif. Les sources de variabilités (variantes de prononciation, phénomènes dits « disfluents » et structures dites «agrammaticales», accents (étrangers, régionaux etc., bruits environnants etc.) sont particulièrement présentes dans ces données. Le travail envisagé portera sur: 1. la modélisation sémantique de l'information à extraire. (i) Un premier volet portera sur l'indexation thématique en fonction des données exploitées et des besoins opérationnels (ex: repérer les conversations où les clients abordent un thème donné). (ii) Un deuxième volet de recherche visera à expliciter les raisons de satisfaction ou d'insatisfaction. Le travail effectué sera examiné dans le cadre des recherches sur la fouille d'opinions (opinion mining). 2. l'adaptation aux transcriptions. Il s'agira d'adapter les techniques d'extraction d'information aux spécificités de l'oral: disfluences, faible cohérence syntaxique, marqueurs pragmatiques (ex: marqueurs de discours, indices dialogiques). Des traitements spécifiques seront à envisager afin de minimiser l'impact du bruit de transcription automatique (erreurs de transcription). Une étape préalable consistera à faire une analyse systématique des erreurs de transcription automatique en fonction de l'information recherchée et du cadre dialogique (profils conseiller/client). Profil recherché : Bac+5 dans le domaine du TALN et/ou du traitement de la parole. Français et Anglais courant. Contexte : Cette thèse s'effectuera entre le LIMSI, groupe TLP (Traitement du Langage Parlé) et le département ICAME (Innovation Commerciale pour l'Analyse des Marchés et de leur Environnement) d'EDF R&D Candidature : Envoyer un CV, une lettre de motivation, et les coordonnées d'au moins un référent universitaire à l'ensemble des adresses suivantes : chloe.clavel@edf.fr , anne.peradotto@edf.fr madda@limsi.fr, ioana@limsi.fr, sophie.rosset@limsi.fr
| |||||||||||||||||||||||||||||
6-5 | (2010-12-17) PhD at INRS Montreal Canada A PhD Position is available at the Multimodal Signal Analysis and Enhancement Laboratory (MuSAE Lab), Institut National de la Recherche Scientifique (INRS-EMT), located in Montreal, Canada. The MuSAE Lab has an opening for a PhD student (full financial support) to work in the area of analysis, assessment, and/or enhancement of speech/audio/biomedical signals. Strong knowledge of digital signal processing (speech/audio/biomedical), pattern recognition, and sensor fusion is required. Strong programming skills in Matlab is preferred. Interested candidates should send their CV's, research interests and relevant publications (if applicable) to Dr. Tiago H. Falk (falk [at] emt [dot] inrs [dot] ca). Note that only short-listed candidates will be contacted. Created in 1969, INRS is a university dedicated to research and graduate studies. At INRS, research is carried out in a context of innovation, action, and partnership with the industry and government. The Energy, Materials, and Telecommunications (INRS-EMT) Research Center focuses its research activities in strategic sectors of communications, photonics and energy. INRS-EMT is located in Montreal, Quebec, Canada, a (bilingual, English/French) city renowned for its great pool of ethnic diversity and its vibrant cosmopolitan ambiance.
| |||||||||||||||||||||||||||||
6-6 | (2010-12-20) Tenure-track position in signal processing and tenure-track position in acoustics at the Aalto University (Espoo, Finland).
Tenure-track position in signal processing and tenure-track position in acoustics at the Aalto University (Espoo, Finland). We invite applications for two tenure-track positions (one in signal processing, one in acoustics) at the Aalto University, Espoo, Finland (formerly known as the Helsinki University of Technology). The tenure track is open to talented individuals who have excellent potential for a scientific career. On the basis of their experience and competence, applicants will be placed on any of the four levels of the tenure track system: Assistant Professor (1), Assistant Professor (2), Associate Professor, or Full Professor. The dead line of the call is January 30, 2011. Further information and instructions are available here: http://www.aalto.fi/en/current/jobs/
| |||||||||||||||||||||||||||||
6-7 | (2011-01-15) 3 PhD positions at Tilburg University (The Netherlands) The Department of Communication and Information Science of the School of Humanities at
| |||||||||||||||||||||||||||||
6-8 | (2011-01-15) 3 Engineer positions open at SYNVO Company Profile Software Engineer for Text-to-Speech • Start Date: At earliest availability Software Architect for Text-to-Speech Tasks and Responsibilities
Start Date: At earliest availability Senior Researcher in Signal Processing Tasks and Responsibilities
• Start Date: At earliest availability Please send your application to Harald Romsdorfer (romsdorfer@synvo.com)
| |||||||||||||||||||||||||||||
6-9 | (2011-01-20) Poste de Professeur, Grenoble INP (ENSIMAG ), Laboratoire LIG France Titre : Job: Poste de Professeur 27, Grenoble INP (ENSIMAG ), Laboratoire LIG
Un poste de professeur, section 27 est à pourvoir pour septembre 2011 à l'ENSIMAG (Grenoble INP) et au laboratoire LIG pour lequel l'activité de recherche concerne notamment le traitement automatique des langues naturelles.
* ENSEIGNEMENT
Composante : ENSIMAG
Profil :
La personne recrutée devra posséder de solides connaissances en informatique et en développement de logiciels. Elle sera amenée à participer aux enseignements scientifiques de tronc commun (algorithmique, bases de données, programmation, théorie des langages, statistique, etc.) et à travailler dans le cadre des enseignements par projet de l'Ensimag. Elle .devra également s’impliquer dans la vie de l’école et de l’établissement par la prise de responsabilités.
Contacts : Brigitte Plateau ( Brigitte.Plateau@grenoble-inp.fr) pour l'ENSIMAG et Yves Maréchal ( Yves.Marechal@g2elab.grenoble-inp.fr) pour Grenoble INP
* RECHERCHE
Profil :
Le laboratoire recherche des candidatures d’excellence permettant de renforcer son encadrement et de participer à sa dynamique. Les 2 thématiques du profil, citées par ordre alphabétique, se déclinent de manière large :
– Accès à l’information : fouille de données massives et complexes, ingénierie des connaissances, recherche d’information, apprentissage automatique, accès pervasif et mobile, etc.
– Interaction : conception d’interfaces homme-machine, dialogue homme-machine. multi-modalité, systèmes perceptifs, mondes virtuels, etc.
Le candidat devra s’intégrer dans l’une quelconque des équipes du LIG, avec un projet de recherche qui développe l’un des aspects du profil du poste, en liaison avec le projet quadriennal « informatique ambiante et durable » du LIG. * AUTRES ACTIVITES Grenoble INP, dans le cadre des Investissements d'avenir est le porteur d'un projet d'Institut autour des énergies décarbonées intelligentes et souhaite que le candidat s'investisse pour son activité de recherche dans cette initiative.
| |||||||||||||||||||||||||||||
6-10 | (2011-02-01) Maitre de conferences UMR CNRS Laboratoire des Sciences de l'Information & des Systèmes à Toulon L'UMR CNRS Laboratoire des Sciences de l'Information & des Systèmes à
| |||||||||||||||||||||||||||||
6-11 | (2011-02-15) Maitre de conferences à l'Université du Mans France Un poste de McF (section 27) est ouvert au concours à l'université du Mans. La personne recrutée fera sa recherche dans l'une des équipes du LIUM : LST ou ingénierie des EIAH.
Les thèmes de recherche ouverts pour ce recrutement dans l'équipe LST sont : traitement automatique de la parole (transcription, interprétation, indexation) et traduction automatique par méthodes statistiques. Contact EIAH Christophe. Choquet at univ-lemans.fr
Contact LST Paul.Deleglise at univ-lemans.fr
| |||||||||||||||||||||||||||||
6-12 | (2011-02-16) Research Associate Position/ Dialogue Systems Group, Cambridge University Engineering Department Research Associate Position
| |||||||||||||||||||||||||||||
6-13 | (2011-02-18) Post docs at Tsinghua University, China Post-doctoral positions
Graduate School at Shenzhen, Tsinghua University, China seeks qualified candidates for several postdoctoral researcher positions:
1. Computational Auditory Scene Analysis (CASA), Active noise control;
2. Speech enhancement, Single-/Multi-channel robust speech recognition;
3. Model based speech synthesis.
The appointments are for two years with possibility of renewal. Salaries are competitive.
Job requirements:
- Applicants are expected to have finished, or be about to finish their Ph.D. degrees, to have a strong background in the area of machine learning and digital signal processing, and to have a track record of publications in top conferences and journals.
- Extensive experience with Matlab and C/C++ (Python a plus).
- Good English communication skills;
Salary:
- Graduate School at Shenzhen, Tsinghua University is located in Shenzhen city, situated immediately north of Hong Kong. Owing to China's economic liberalization, the area became China's first and the most successful Special Economic Zones.
- Salaries are in the range 20,000 to 25,000 USD per year based on the candidates' experience. An apartment with about 70 square meters (two bed rooms, one living room and one kitchen room)is provided inside the campus and the rent is less than 200 USD per month.
- The laboratory offers a creative international environment, and works on an interdisciplinary team (with vision processing, active HCI, embedded system, etc.) We provide a possibility to conduct competitive research on a global scale. There will be ample opportunities to cooperate with some of the best groups in Hong Kong, Europe and the USA.
Application Instructions:
Interested applicants should send a brief statement of research interest and a CV to Prof. Weifeng Li (li.weifeng at sz.tsinghua.edu.cn) Tel: +86 755 2603 6023 Address: Room 103A, F Building, Tsinghua Campus, Nanshan District, 518055 Shenzhen, P.R. China
| |||||||||||||||||||||||||||||
6-14 | (2011-02-22) Programming position at ELDA Paris France ELDA offers a position for its Language Resource Production and Evaluation activities working in the framework of European projects.
The position is related to a number of NLP activities within ELDA, with a focus on the development of web-service architectures for the automatic production and distribution of language resources. The candidate may also be involved in the creation of LR repositories, NLP applications development and/or evaluation, etc.
| |||||||||||||||||||||||||||||
6-15 | (2011-02-24) Research position at NICT Japan
Spoken Language Communication Group
| |||||||||||||||||||||||||||||
6-16 | (2011-03-01) Post-doctoral position in Automatic Speech Recognition at IRISA, France.Post-doctoral position in Automatic Speech Recognition at IRISA, France. In the framework of the project 'Automatic System Harnessing' (ASH), funded by the French National Research Agency, IRISA seeks to hire a postdoctoral researcher in the field of automatic speech recognition. Landmark-driven paradigms for large vocabulary speech recognition will be the main topic of interest [1,2,3], with a focus on developing bidirectional interaction mechanisms between landmark detection and speech transcription. The ASH project is a national project aiming at developing collaborative approaches to speech recognition by tight coupling of several ASR systems or by coupling an ASR system with landmark detectors. The idea of coupling (harnessing) systems originates from the driven decoding principle as explained in [4]. The project is leaded by the Laboratoire d'Informatique de l'Université du Mans (LIUM) with the Laboratoire d'Informatique d'Avignon (LIA) as partner in addition to IRISA. The postdoc will be in charge of designing and implementing interaction mechanisms between the IRISA's ASR system and that of the other partners, during the decoding stage. Similar mechanims will be investigated for joint ASR and landmark detection in Irisa's ASR system. These research activities will be incoporated in fit in the framework of the upcoming national benchmark ETAPE/ESTER 3. Candidate should be a recent PhD in computer science or electrical engineering with strong experience in large vocabulary speech recognition (preferably search or acoustic modeling) and/or audio signal processing and classification. Candidates should have good practical and theoretical knowledge in pattern recognition. We are primarily interested in statistical approaches to pattern recognition but skills in discriminative techniques, in particular neural networks, are welcome. As we are mostly dealing with French data, knowledge of the French language is a plus but not required. IRISA is a French public laboratory located in Rennes (France), affiliated to CNRS, the French national Center for Scientific Research. The candidate will join the leading multimedia group TEXMEX at IRISA (http://www.irisa.fr) and develop his research in strong collaboration with the speech and audio processing group METISS. He will be part of the ASR&NLP workgroup which gathers leading scientists in speech recognition, natural language processing and information retrieval. Work will be carried out in close collaboration with a PhD student on landmark-driven speech recognition and in interaction with ASH partners. The position, to be fulfilled as soon as possible, is opened for one year, possibly with a 6 months extension. Gross salary is 2500 euros per month (2040 net pay) and includes social benefits. Please contact Guillaume Gravier (guig@irisa.fr) for more details or send an extended resume for application. [1] M. Hasegawa-Johnson, S. Borys and K. Chen, Experiments in Landmark-Based Speech Recognition. Sound to Sense: Workshop in Honor of Kenneth N. Stevens, June, 2004. [2] G. Gravier and D. Moraru. Towards phonetically-driven hidden Markov models: Can we incorporate phonetic landmarks in HMM-based ASR? In Proc. ISCA Tutorial and Research Workshop on Non Linear Speech Processing (NOLISP), 2007. [3] S. Sainath. Island driven search using broad phonetic classes. IEEE Worskhop on Automatic Speech Recognition and Understanding, 2009. [4] B. Lecouteux, G. Linarès, Y. Estève, and G. Gravier. Generalized driven decoding for speech recognition system combination. In IEEE Intl. Conf. on Acoustics, Speech and Signal Processing, 2008.
| |||||||||||||||||||||||||||||
6-17 | (2011-03-18) Ingenieur R/D INRIA Rennes France Le centre INRIA de Rennes Bretagne Atlantique recrute en CDD :
| |||||||||||||||||||||||||||||
6-18 | (2011-03-09) Ingenieur R/D INRIA Rennes (France) Le centre INRIA de Rennes Bretagne Atlantique recrute en CDD :
| |||||||||||||||||||||||||||||
6-19 | (2011-03-09)Internship for Web Interface for natual language data bases INRIA Paris Stage en développement web interface graphique
Recherche stagiaires en ingéniérie Web pour le développement d?un frontal web dans le cadre d?un projet CNRS/ANR en traitement automatique des langues naturelles (http://rhapsodie.risc.cnrs.fr/fr/index.html, laboratoires partenaires IRCAM, Paris ; MODYCO, Nanterre)
Objectifs : Développement d?une interface web de consultation, visualisation et d?annotation de données orales structurées (3h d'enregistrement audio avec la transcription et des analyses linguistiques)
Tâche demandée : Mise en place d?une interface web permettant d?accéder à des données audio et à des annotations sur ces données par l?intermédiaire d?un langage de requête constitué au préalable. Cette interface proposera : - des formulaires de recherche s?appuyant sur un moteur existant - la représentation graphique ergonomique des annotations à l'aide de différentes visualisations des données audio et textuelles - la possibilité de réviser les annotations fournies - des statistiques sur les données et leurs représentations graphiques - un backoffice d?administration des données et des utilisateurs
Compétence techniques requises :
- Serveur : PHP ou un autre langage script (Python, Perl) et MySQL - Client Web : Javascript, HTML La connaissance de AJAX, Json, JQuery et SVG ainsi qu'en développement des outils Web avec des accès réservés sera un plus. Un intérêt pour les structures de la langue sera apprécié mais aucune connaissance particulière en parole et linguistique n?est requise. Le candidat collaborera avec l?équipe du laboratoire (linguistes et informaticiens) pour formaliser les spécifications fonctionnelles de l?application.
Indéminités mensuelles de stage. 417,09 euros
Lieu de travail : laboratoire MODYCO, Université de paris Ouest Nanterre, télétravail possible
avec mise à disposition d?ordinateur.
Début du stage : avril 2011 ou le plus tôt possible.
Durée du stage : 3 mois minimum et possibilité d?enchaîner sur un CDD de 6 à 12 mois. Contacter : Atanas Tchobanov atanas@u-paris10.fr et Kim Gerdes kim@gerdes.fr
Nicolas OBIN: PhD Student at IRCAM Analysis/Synthesis Team -
contact | nobin@ircam.fr -
? : 33 (0) 1 44 78 48 90, Fax : 33 (0)1 44 78 15 40
| |||||||||||||||||||||||||||||
6-20 | (2011-03-15) Engineer in Multimedia TCL AMERICATCL Research America (www.TCL-AMERICA.com<http://www.tcl-america.com/>), the research arm of TCL Corporation in North America, is currently recruiting Research Scientists and Architects to conduct R&D efforts in the fields of multimedia, communications and smart computing. We offer competitive compensation, benefits and bonus to our employees. Established in 1981, TCL Corporation<http://www.tcl.com/en.php> is a global leader in consumer electronics and communications with over 50,000 employees worldwide. TCL runs the world 6th largest TV business and 7th largest mobile phone business with globally respected brands Alcatel, Thomson and TCL. In the North America market, TCL ships Alcatel mobile phones<http://www.alcatel-mobilephones.com/> and TCL TVs<http://www.tclusa.com/>. More information of the company can be found at www.TCL-AMERICA.com<http://www.tcl-america.com/> List of Openings: (http://www.tcl-america.com/careers.html) Senior Smart TV Architect Senior Multimedia Architect Research Scientist Visiting Professor Visiting Scientist Please submit your resume in pdf to CAREER.AMERICA@TCL.COM<mailto:CAREER.AMERICA@TCL.COM>. Our human resource staff will follow up with your application shortly. TCL Research America http://www.tcl-america.com<http://www.tcl-america.com/>
| |||||||||||||||||||||||||||||
6-21 | (2011-03-22) Thesis in psycholinguistic at the University of Geneva Switzerland Le laboratoire de psycholinguistique expérimentale de l'Université de Genève met au concours:
1 poste de doctorant/e FNS en psycholinguistique expérimentale à 100% Description: Ce poste s'inscrit dans le cadre du projet de recherche financé par le Fonds National Suisse de la Recherche Scientifique (FNS) dont l’intitulé est le suivant : « Relative contribution of multiple segmentation cues to lexical identification and acquisition ». L’objectif principal du projet est d’explorer comment différents indices sous-lexicaux fournis par la structure phonétique, phonologique et rythmique de la parole contribuent à l’identification et à l’acquisition lexicale. Le/la candidat/e participera à la récolte et à l’analyse des données et développera un projet de thèse sur l'un des thèmes abordés par la recherche. Profil: Maîtrise en Psychologie, Master en Sciences du Langage avec spécialisation phonétique/phonologie (ou titre jugé équivalent). Les travaux antérieurs du/de la candidat/e devront mettre en évidence un intérêt particulier pour la psycholinguistique expérimentale. La motivation à faire une thèse de doctorat dans le cadre de la recherche soutenue par le projet sera déterminante dans le choix du/de la candidat/e. Entrée en fonction: juillet 2011 (ou à convenir) Durée du projet: 3 ans Lieu de travail: laboratoire de psycholinguistique expérimentale, Université de Genève Salaire brut: de CHF 40’200 à CHF 46’200 Délai pour le dépôt des candidatures: 15 Mai 2011 (prévoir un entretien éventuel début juin) Dossier de candidature: Les personnes intéressées sont priées de soumettre un dossier de candidature complet, incluant une lettre de motivation, leur Curriculum Vitae ainsi que la copie électronique d'une publication scientifique et/ou du mémoire de fin d'études (ou, à défaut, une recommandation du directeur/trice de mémoire). Pour de plus amples informations, veuillez vous adresser à Odile Bagou : Odile.Bagou@unige.ch. Le dossier doit être envoyé par e-mail à l'attention du Prof. Ulrich Frauenfelder (Ulrich.Frauenfelder@unige.ch).
| |||||||||||||||||||||||||||||
6-22 | (2011-03-30) Research associate position in the Dialog Group at Cambridge Univ. UKResearch Associate, Dialogue Systems Group Applications are invited for a Research Associate position, working on automatic speech recognition. The project will be based in the Dialogue Systems Group, in the machine intelligence lab at the Cambridge University Engineering Department with an anticipated start date of May 1st, 2011 or soon after. The aim of the project is to build a server-based robust spoken dialogue system providing tourist information for in-car users. The research associate will be involved in building the system's real-time conversational speech recognition engine. Tasks will include acoustic/language model training, implementing noise robustness techniques, decoder optimization, silence detection and turn taking research. ATK and HTK will be used as the speech recognition platform. The successful candidate will have a PhD in a relevant area, such as speech recognition or computer science, and be familiar with HMM based speech recognition techniques. Preference will be given to candidates with experience in noise robustness research, end-to-end implementation of speech recognition systems, and those familiar with the ATK/HTK platforms. The Dialogue Systems Group is a world-leading research group working on robust statistical spoken dialogue systems. During the project, there will be good opportunities to attend international conferences and workshops. Interested candidates should send their CV as soon as possible to Dr Kai Yu (ky219@cam.ac.uk) or Prof. Steve Young (sjy@eng.cam.ac.uk). Salary will be in the range £27319 to £35646 and the closing date is 14th April 2011. More information on the Dialogue Systems Group can be found at http://mi.eng.cam.ac.uk/research/dialogue.
| |||||||||||||||||||||||||||||
6-23 | (2011-04-01) Research Position in Speech Recognition for Underresourced Languages at K.U.Leuven - ESAT, Belgium
Today's speech recognition systems require hundreds of hours of example data for training the acoustic models. While such large corpora are available for the major languages, this is not the case for smaller languages, making them 'under-resourced'�. One of the underlying reasons for this data hungriness is that the dimensionality of feature vectors used in state-of-the-art speech recognition systems (typically in the range 30-40) is much larger than the intrinsic dimensionality of speech which is estimated to be 7-10 only. Efforts to make the intrinsic dimensionality smaller have been largely futile as the constraints are too complex for our by and large linear techniques. This inefficiency in basic representation, combined with other inefficiencies in mainstream context-dependent modeling makes that the hundreds of thousands parameters that constitute an acoustic model are largely redundant. The objective of this project is to apply novel mathematical techniques (e.g. spectral clustering) that can capture constraints - not in the feature space - but in the model space, i.e. in the underlying HMM parameters. Such constraints will lead to lesser requirements on the size of the training databases and should increase robustness in all situations where we don't have large corpora available, such as speaker adaptation, accent adaptation or modeling of under-resourced languages. Apart from general principles, two test cases will be be studied in more detail : i) 'Afrikaans'�, for which data from Dutch and Flemish can be reused; ii) languages form the Bantu family as spoken in South Africa for which we can only bootstrap from a wide set of rather unrelated languages. Qualifications Candidates ideally have a university degree in engineering, computer science or applied mathematics. Skills and experience in any of the following areas are welcomed: Position Within this project there is funding for a 4yr Ph.D. scholarship. Alternatively we will also accept applications for a 2 yr junior post-doc with significant relevant experience. Contact Dirk Van Compernolle - compi@esat.kuleuven.be Project Partners This research will be carried out the K.U.Leuven, Belgium in the context of the AMODA project in collaboration with Council for Scientific and Industrial Research (CSIR), Pretoria, South Africa.
| |||||||||||||||||||||||||||||
6-24 | (2011-05-04) Post-doctoral researcher, audio signal processing, at Microsoft Research Post-doctoral researcher, audio signal processing, at Microsoft Research
Position Title: Post-doctoral researcher Organization Name: Microsoft Research Location: Redmond, WA, USA.
Microsoft Research (http://research.microsoft.com/en-us/) is looking for one or more post-doctoral researchers in the general area of audio signal processing. They will perform research in sound capture and rendering, speech enhancement, arrays of transducers processing, psychoacoustics, human perception of sounds and images.
Strong background in the area of audio and speech signal processing, good publication record, and experience in creating a working audio systems are requirements. Expertise in speech enhancement and/or psychoacoustics is a strong advantage; good engineering and coding skills are preferred. Ability or experience to work with engineering teams is a plus.
For enquiries please send e-mail to Dr. Ivan Tashev at ivantash@microsoft.com.
| |||||||||||||||||||||||||||||
6-25 | (2011-05-07) Eight Research Positions in Speech Synthesis and Speech Recognition at Universities of Edinburgh, Cambridge, SheffieldEight Research Positions in Speech Synthesis and Speech Recognition at Universities of Edinburgh, Cambridge, SheffieldNATURAL SPEECH TECHNOLOGY - http://www.natural-speech-technology.org
|