IEEE SPEECH TECHNICAL COMMITTEE NEWSLETTER

December, 2005

INTRODUCTION:

Welcome to the IEEE Signal Processing Society Speech Technical Committee (STC) newsletter. Contributions of events, publications, workshops, and career information to the newsletter are welcome. Please send to the new STC Newsletter Editor Mike Seltzer (mseltzer_@_microsoft_com) or Rick Rose (rose_@_ece_mcgill_ca). Archives of recent STC Newsletters can be found on the STC website.

STC NEWS:
ICASSP2006 Paper Review Process (Mazin Rahim)

SPECIAL ISSUES OF TRANSACTIONS:
Special Issue of the IEEE Transactions on SAP: Objective Quality Assessment of Speech and Audio

NEW WORKSHOP ANNOUNCEMENTS:
Tutorial and Research Workshop on Speech Recognition and Intrinsic Variation
HLT-NAACL 2006 Call for Papers
HLT-NAACL 2006 Call for Demonstrations
2006 ISCA Tutorial and Research Workshop on Statistical and Perceptual Audition (SAPA2006)
2006 ELRA 5th Conference on Language Resources and Evaluation (LREC2006)
IASTED 2006 International Conference on signal processing, pattern recognition, and applications (SPPRA 2006)

CAREERS:
Position Available: Post Doc / Visiting Scientist at University of Washington
Transitions: ASR Researchers Take New Positions

LINKS TO WORKSHOPS AND CONFERENCES:
Links to conferences and workshops organized by date (Rick Rose)

back to top

---

ICASSP2006 Paper Review Process

Speech and Language Technical Committee received a record high of 668 papers for ICASSP 2006 not including the special session papers. This is a historical record high in the speech and language area for this conference, and amounts to over 18% increase from last year. The majority of this increase is attributed to more papers on Robustness, Speech Recognition and Synthesis, and Language Processing. Nearly all papers received three reviews from experts in this community. The review committee involved 10 Area Chairs, 48 Technical Committee members and 180 external reviewers. A list of the reviewers can be found below. We would like to thank all those who made contribution to the review process in the tight schedule.

• List of Area Chairs
• List of STC Team Members
• List of All Reviewers

The accept rate for this year is about 46% which is less than the 50% we had last year. We will be having 13 lecture sessions and 21 poster sessions, in addition to 3 special sessions and 3 other tutorials in the areas of speech and language processing. See the ICASSP 2006 website for more details, and the notification of individual paper acceptance will be delivered on January 9th.
We are look forward to seeing you in Toulouse!
back to top

---

Call for Papers

Special Issue of
The IEEE Transactions on Speech and Audio Processing
Objective Quality Assessment of Speech and Audio

Objective Quality Assessment of Speech and Audio is an interdisciplinary research area to build computational models that aim to achieve human performance in quality estimation of speech and audio. The estimation of quality is becoming more important especially in telecommunication applications, where Quality of Service is one of the key considerations.

The goal of this special issue is to present recent progress and advances in this area as well as remaining challenges. We invite original, previously unpublished research works in all areas relevant to the field. In particular, paper submissions are encouraged on the following topics:

• Subjective basis for objective quality assessment
• Waveform models - based on waveforms of speech and audio
• Parametric models - based on telecommunication or broadcast network parameters
• Intrusive models
• Non-intrusive (single-ended or output-based) models
• Objective diagnosis of quality impairment
• Objective and subjective assessment of conversational quality
• Issues and applications relevant to real-world problems

Submission procedure:

Prospective authors should prepare manuscripts according to the Information for Authors as published in any recent issue of the Transactions and as available on the web at http://www.ieee.org/organizations/society/sp/infotsa.html. Note that all rules will apply with regard to submission lengths, mandatory overlength page charges, and color charges.

Manuscripts should be submitted electronically through the online IEEE manuscript submission system at http://sps-ieee.manuscriptcentral.com/. When selecting a manuscript type, authors must click on "Special Issue of T-SA on Objective Quality Assessment of Speech and Audio." Authors should follow the instructions for the IEEE Transactions on Speech and Audio Processing and indicate in the Comments to the Editor-in-Chief that the manuscript is submitted for publication in the Special Issue on Objective Quality Assessment of Speech and Audio. We require a completed copyright form to be signed and faxed to 1-732-562-8905 at the time of submission. Please indicate the manuscript number on the top of the page.

Schedule:

Guest Editors: back to top

---

Call for papers

ISCA Tutorial and Research Workshop (ITRW) on Speech Recognition and Intrinsic Variation

May 20, 2006 - Toulouse (France)

Introduction

Major progress is being recorded regularly on both the technology and exploitation of Automatic Speech Recognition (ASR) and spoken language systems. Flexible solutions and user satisfaction under some circumstances are still challenging topics however. Particularly, current research is emphasizing deficiencies in dealing with speech intrinsic variations. For instance, specific aspects of the speech signal like foreign accents, precludes the use by specific populations. Also, some applications, like directory assistance, particularly stress the core recognition technology due to the very high active vocabulary (application perplexity). Besides, speech models seem to remain inadequate for handling unconstrained spontaneous speech and dialogs. Investigations on factors affecting speech realization and variations within the speech signal that make the ASR task challenging are hence very valuable for future developments of spoken language systems.

The goal of this workshop is to strengthen common understanding and research trends that deal with these topics. Original studies and algorithms related to ASR sensitivity, genericity, user-independence, robustness or adaptation to variations in speech should be covered. Besides, original contributions from the fields of phonetic science, as well as human speech perception and recognition are warmly encouraged. Finally, we also invite submission of papers about application and services scenarios involving specific speech variations.

Topics

Papers with an emphasis placed on the following topics are encouraged:

• Speech variability, including foreign and/or regional accented speech modeling and recognition,
• Children and/or elderly speech modeling and recognition,
• Speech non-stationarity and relevant analysis methods,
• Speech spectral (e.g. vocal tract length) and temporal (e.g. speaking rate) variations,
• Spontaneous speech modeling and recognition,
• Speech variation due to emotions, fatigue, frustration,...
• Speech corpora covering sources of variation,
• Modeling of speech variability,
• Acoustic-phonetic correlates of variations,
• Impact and characterization of speech variations on ASR analysis, models and performance,
• Speaker adaptation and normalization (e.g., VTLN); Speaker adapted training methods,
• Novel general or specific compensation and adaptation algorithms,
• Novel analysis and modeling structures and algorithms for handling speech variations,
• Alternative Statistical and Machine Learning Methods for General ASR (e.g., no-HMM methods),
• Man/machine confrontation; ASR and HSR (human speech recognition) interdisciplinary views,
• Experiments making use of the OLLO speech corpus (http://www.divines-project.org/),
• Intrinsic variations in multimodal recognition,
• Review papers on these topics are also welcome.

Submission

Prospective authors are invited to submit full-length (4-6 pages) previously unpublished papers via the workshop website, in the submission section. Guidelines for paper submission are provided there. The full-length papers will be selected through a peer review process.

The proceedings will be available via the workshop web site.

Structure

The workshop will consist of oral and poster sessions, as well as talks by invited speakers. The workshop will be limited to 100 participants and about 50 presentations.

Venue

This event is organized as a satellite of the ICASSP 2006 conference, which is held in Toulouse, France in May 2006. The workshop will also take place in Toulouse, on 20 May 2006, just after the conference, which ends May 19. More information on the venue will be provided on the workshop website.

Important Dates

Submission deadline: February 1, 2006
Notification of acceptance: March 1, 2006
Final manuscript due: March 15, 2006
Program available: March 22, 2006
Workshop registration deadline: March 29, 2006
Workshop: May 20, 2006 (after ICASSP 2006)

Commitee

Pierre-Albert Breton, Thales Avionics, France
Renato De Mori, Université d'Avignon, France
Olivier Deroo, Acapela Group, Belgium
Stéphane Dupont, Multitel, Belgium
Luciano Fissore, Loquendo, Italy
Roberto Gemello, Loquendo, Italy
Alexandre Girardi, Multitel, Belgium
Denis Jouvet, France Telecom, France
Katrin Kirchhoff, U. Washington, WA, USA
Birger Kollmeier U. Oldenburg, Germany
Pietro Laface, Politecnico di Torino, Italy
Marco Matassoni, ITC-IRST, Italy
Alfred Mertins, U. Oldenburg, Germany
Alexandros Potamianos, Technical university of Crete, Greece
Christophe Ris, Multitel, Belgium
Richard Rose, McGill University, Canada
José C. Segura, Universidad de Granada, Spain
Piergiorgio Svaizer, ITC-IRST, Italy
Louis ten Bosch, Radboud University, Nijmegen, The Netherlands
Christian Wellekens, Eurecom, France
back to top

---

HLT-NAACL 2006 continues the combination of the Human Language Technology Conferences (HLT) and North American Chapter of the Association for Computational Linguistics (NAACL) Annual Meetings begun in 2003. Human language technology incorporates a broad spectrum of disciplines working towards enabling computers to interact with humans using natural language, and providing services such as speech recognition, automatic translation, information retrieval, text summarization, and information extraction.

HLT-NAACL 2006 will run from Sunday June 4 through Friday June 9. The schedule will include full papers, late-breaking (short) papers, demonstrations, as well as pre- and post-conference tutorials and workshops. The conference organization is overseen by a board representing the North American Chapter of the Association for Computational Linguistics (NAACL), HLT funding agencies in North America, as well as the SIGIR and ISCA communities.

Topics of Interest

The conference invites the submission of papers on substantial, original, and unpublished research on all aspects of human language processing, with special interest in synergistic combinations of language technologies (e.g., Speech with Information Retrieval, Machine Translation with Speech, Question Answering with Natural Language Processing, etc.). Topics of interest include but are not limited to:

• Speech processing, including:
  
  • Speech recognition and speech generation
  • Rich transcription: automatic annotation of information structure and sources in speech
• Information extraction, text summarization, and question answering
• Information retrieval
• Computational analysis of phonology, morphology, syntax, semantics, pragmatics, discourse, style
• Statistical and learning techniques for language processing, including
  • Corpus-based language modeling
  • Lexical and knowledge acquisition
• Language generation and text planning
• Multilingual processing, including
  • Machine translation of speech and text
  • Cross-language information retrieval
  • Multi-lingual speech recognition and language identification
• Multimodal representations and processing
• Evaluation, including
  • Glass-box evaluation of HLT systems and system components
  • Black-box evaluation of HLT systems in application settings
• Development of language resources, including
  • Lexicons and ontologies
  • Treebanks, proposition banks, and frame banks
• Understanding of human communication, including
  • Natural language interfaces
  • Dialogue structure and dialogue systems
  • Message and narrative understanding system
  
  

---

SUBMISSION INFORMATION

Full Papers

Requirements: Submissions must describe original, completed, unpublished work, and include concrete evaluation results when appropriate. Submissions will be judged on correctness, originality, technical strength, significance and relevance to the conference, and interest to the attendees. As reviewing will be blind, no information identifying the authors should be in the paper: this includes not only the authors' names and affiliations, but also self-references that reveal authors' identities; for example, "We have previously shown (Smith 1999)" should be changed to "Smith (1999) has previously shown". Separate identification information is required, and will be part of the web submission process.

Format: Submissions must be electronic in PDF, should follow the two-column format of ACL proceedings, and should not exceed eight (8) pages, including references. Please see the conference website for detailed typesetting specifications. Authors are strongly encouraged to use the LaTeX or Microsoft Word style files available on the conference website.

Reviewing: The reviewing of the papers will be blind. Reviewing will be managed by a Conference Program Committee consisting of senior Program Committee Members and associated Program Committee Members. Each submission will be reviewed by at least three program committee members.

Submission procedure: A PDF file of the paper must be uploaded onto the system by 11:59pm EST of the deadline. Papers submitted after that time will not be reviewed. Authors who cannot submit a PDF file electronically should contact the program co-chairs
(, , or ) before the due date to work out alternate arrangements.

Late-Breaking (Short) Papers

The procedure for Short Papers submissions is identical to that of the Full Papers, with the following differences:

1. They may be accepted for oral presentation in plenary OR for presentation in a poster session;
2. The deadlines are later for short papers and posters than for full papers;
3. Short papers are restricted to four (4) pages in length, using the two-column ACL format;
4. Only two reviews per submission are guaranteed.

General Conference Chair: Robert Moore (Microsoft Research)

Program Co-Chairs:

Jeff Bilmes (University of Washington)

Jennifer Chu-Carroll (IBM T.J. Watson Research Center)

Mark Sanderson (Sheffield University)

Senior Program Committee Members:

Johan Bos (University of Edinburgh) Dragomir Radev (University of Michigan)

Jamie Callan (CMU) Owen Rambow (Columbia University)

Joyce Chai (Michigan State University) Steve Renals (University of Edinburgh)

Jason Eisner (Johns Hopkins University) Stefan Riezler (PARC)

Mark Gales (University of Cambridge) Amanda Stent (SUNY Stony Brook)

Fred Gey (Berkeley) Rohini Srihari (SUNY Buffalo)

Roxana Girju (UIUC) Michael Strube (EML Research)

Mark Hasegawa-Johnson (UIUC) Christoph Tillmann (IBM Watson)

Julia Hirschberg (Columbia University) Peter Turney (National Research Council Canada)

Alon Lavie (CMU) Ellen Voorhees (NIST)

Wei Ying Ma (Microsoft Beijing) Ralph Weischedel (BBN)

Mehryar Mohri (NYU) Fei Xia (University of Washington)

Marius Pasca (Google) ChengXiang Zhai (UIUC)

Gerald Penn (University of Toronto) Ming Zhou (Microsoft Beijing)

Local Arrangements Chair: Satoshi Sekine (New York University)

Important Dates

All submissions or camera-ready copies are due by 11:59pm EST on the date specified above.

Conference Venue

The conference will be held at the New York Marriott at the Brooklyn Bridge, which is located just three subway stops from Downtown and ten stops from Midtown, the centers of the Big Apple.
top of page

---

Proposals are invited for the HLT-NAACL 2006 Demonstrations Program. This program is aimed at offering first-hand experience with new systems, providing opportunities to exchange ideas gained from creating systems, and collecting feedback from expert users. It is primarily intended to encourage the early exhibition of research prototypes, but interesting mature systems are also eligible. Submission of a demonstration proposal on a particular topic does not preclude or require a separate submission of a paper on that topic; it is possible that some but not all of the demonstrations will illustrate concepts that are described in companion papers.

Demo Co-Chairs:
John Dowding, University of California/Santa Cruz
Natasa Milic-Frayling, Microsoft Research, Cambridge, United Kingdom
Alexander Rudnicky, Carnegie Mellon University.

Areas of Interest

We encourage the submission of proposals for demonstrations of software and hardware related to all areas of human language technology. Areas of interest include, but are not limited to, natural language, speech, and text systems for:

- Speech recognition and generation;
- Speech retrieval and summarization;
- Rich transcription of speech;
- Interactive dialogue;
- Information retrieval, filtering, and extraction;
- Document classification, clustering, and summarization;
- Language modeling, text mining, and question answering;
- Machine translation;
- Multilingual and cross-lingual processing;
- Multimodal user interface;
- Mobile language-enabled devices;
- Tools for Ontology, Lexicon, or other NLP resource development;
- Applications in growing domains (web-search, bioinformatics, ...).

Please be referred to the HLT-NAACL 2006 CFP (http://nlp.cs.nyu.edu/hlt-naacl06/cfp.html ) for a more detailed but not necessarily an exhaustive list of relevant topics.

Important Dates
Submission deadline: March 3, 2006
Notification of acceptance: April 6, 2006
Submission of final demo related literature: April 17, 2006
Conference: June 4-9, 2006

Submission Format

A demo proposal should consist of the following parts:

- An extended abstract of up to four pages, including the title, authors, full contact information, and technical content to be demonstrated. It should give an overview of what the demonstration is aimed to achieve, how the demonstration illustrates novel ideas or late-breaking results, and how it relates to other systems or projects described in the context of other research (i.e., references to related literature).

- A detailed requirement description of hardware, software, and network access expected to be provided by the local organizer. Demonstrators are encouraged to be flexible in their requirements (possibly preparing different demos for different logistical situations). Please state what you can bring yourself and what you absolutely must be provided with. We will do our best to provide equipment and resources but at this point we cannot guarantee anything beyond the space and power supply.

- A concise outline of the demo script, including the accompanying narrative, and either a web address to access the demo or visual aids (e.g., screen-shots, snapshots, or sketches). The demo script shouldbe no more than 6 pages.

Procedure

Demo proposals should be submitted electronically to the demo co-chairs at hlt-naacl06-demonstrations@cs.nyu.edu .

Reviewing

Demo proposals will be evaluated on the basis of their relevance to the conference, innovation, scientific contribution, presentation, and usability, as well as potential logistical constraints.

Publication

The accepted demo abstracts will be published in the Companion Volumne to the Proceedings of the HLT-NAACL 2006 Conference.

Further Details

Further details on the date, time, and format of the demonstration session(s) will be determined and provided at a later date. Please send any inquiries to the demo co-chairs at hlt-naacl06-demonstrations@cs.nyu.edu.

Please check http://nlp.cs.nyu.edu/hlt-naacl06/cfp.html for latest updates.
top of page

---

Papers are solicited for the 2006 Workshop on Statistical and Perceptual Audition (SAPA2006), to be held in Pittsburgh PA as a satellite to ICSLP 2006 <http://www.interspeech2006.org/>.

Following on from the successful SAPA2004 workshop <http://www.sapa2004.org> (in Jeju, Korea), the objective of the SAPA2006 workshop is to bring together researchers considering perceptually-motivated problems in sound and speech analysis and understanding, employing statistical and machine learning tools.

There is a wide area of overlap between more heuristic models of human auditory function and purely pattern recognition approaches that are independent of human audition; SAPA aims to be the forum for presentation and discussion of this promising and expanding field.

This will be a one-day workshop with a limited number of oral presentations, chosen for breadth and provocation, and an informal atmosphere to promote discussion. We hope that the participants in the workshop will be exposed to a broader perspective, and that this will help foster new research and interesting variants on current approaches.

Papers describing relevant research and new concepts are solicited on, but not limited to, the following topics:

* Generalized audio analysis
* Speech analysis
* Music analysis
* Audio classificationy
* Scene analysis
* Signal separation
* Speech recognition
* Multi-channel analysis

In all cases, preference will be given to papers that clearly involve both perceptually-defined or perceptually-related problems, and statistical or machine-learning based solutions. Manuscripts must be between 4 and 6 pages long, in standard ICSLP double-column format. Accepted papers will be published in the workshop proceedings.

Papers must be recieved by 21 April 2006 (two weeks after the ICSLP deadline). The results of the paper review will be posted by 9 June 2006 (same as ICSLP).

Organizers:

Dr. Bhiksha Raj
Research Scientist
Mitsubishi Electric Research Labs,
Cambridge, MA, USA, 02139
bhiksha@merl.com
617 621 7593
Dr. Paris Smaragdis
Research Scientist
Mitsubishi Electric Research Labs,
Cambridge, MA, USA, 02139
paris@merl.com
617 621 7561
Prof. Daniel Ellis
Associate Professor
Columbia University
New York
dpwe@ee.columbia.edu
212 854 8928
top of page

---

For more information, please contact:
IASTED
#80 4500 - 16th Avenue N.W.
Calgary, Alberta
Canada T3B 0M6
Tel: 403-288-1195
Fax: 403-247-6851
E-mail: calgary@iasted.org
Website: http://www.iasted.org
top of page

---

5th Conference on Language Resources and Evaluation

CONFERENCE AIMS

In the Information Society, the pervasive character of Human Language Technologies (HLT) and their relevance to practically all fields of Information Society Technologies (IST) has been widely recognised. Two issues are considered particularly relevant : the availability of Language Resources (LRs) and the methods for the evaluation of resources, technologies, products and applications. Substantial mutual benefits are achieved by addressing these issues through international cooperation.

The term language resources refers to sets of language data and descriptions in machine readable form, such as written or spoken corpora and lexica, annotated or not, multimodal resources, grammars, terminology or domain specific databases and dictionaries, ontologies, multimedia databases, etc. LRs also cover basic software tools for their acquisition, preparation, collection, management, customisation and use. LRs are used in many types of components/systems/applications, such as software localisation and language services, language enabled information and communication services, knowledge management, e-commerce, e-publishing, e-learning, e-government, cultural heritage, linguistic studies, etc. This large range of usages makes the LRs infrastructure a strategic part of the e-society, where the creation of a basic set of LRs for all languages must be ensured in order to bring all languages to the same level of usability and availability. The relevance of the evaluation for language technologies development is increasingly recognised. This involves assessing the state-of-the-art for a given technology, measuring the progress achieved within a programme, comparing different approaches to a given problem, assessing the availability of technologies for a given application, product benchmarking, and assessing system usability and user satisfaction.

The aim of the LREC conference is to provide an overview of the state-of-the-art, explore new R&D directions and emerging trends, exchange information regarding LRs and their applications, evaluation methodologies and tools, ongoing and planned activities, industrial uses and needs, requirements coming from the new e-society, both with respect to policy issues and to technological and organisational ones. LREC provides a unique forum for researchers, industrials and funding agencies from across a wide spectrum of areas to discuss problems and opportunities, find new synergies and promote initiatives for international cooperation in the areas mentioned above, in support to investigations in language sciences, progress in language technologies and development of corresponding products, services and applications.

back to top

---

The Signal, Speech and Language Interpretation Lab (SSLI) at the University of Washington (UW) is looking for an entry-level speech scientist in Mandarin speech recognition, which is part of an exciting Mandarin speech to English text translation project.

Term: January-September, 2006 with possible extension
Qualifications:
(1) For the post-doc position, a Ph.D. in either Electrical Engineering or Computer Science is required. For the visiting scientist position, a master degree in either of the above two fields is preferred.
(2) Fluent in both Mandarin and English.
(3) Understanding of state-of-the-art large-vocabulary speech recognition systems.

Job responsibility: Research and develop acoustic and language models and algorithms for Mandarin broadcast news speech recognition.

Applications should include a vita and the names of at least 2 references (with both a telephone number and email address), sent to Mei-Yuh Hwang at mhwang@ee.washington.edu or via postal mail to

Dr. Mei-Yuh Hwang
Senior Research Scientist
Dept. of Electrical Engineering
PO Box 352500
Seattle, WA 98195-2500

UW is an Affirmative Action/Equal Opportunity Employer. Applications from women and minorities are especially encouraged. Hiring is contingent on eligibility to work in the United States. For more information about SSLI, please visit http://ssli.ee.washington.edu/.
back to top

---

ASR Researchers Take New Positions

The STC Newsletter would like to provide announcements of professors, researchers, and developers in the speech area taking new positions. If you have moved lately or are in the process of moving to a new position in the new future, send your new contact information to the STC Newsletter so it can be posted in the next edition.

• In August of 2005, Giusseppi Riccardi left AT&T Labs - Research and joined the Engineering Faculty of University of Trento Italy with appointmentto the Department of Information and Communications (www.dit.unitn.it). He will be heading the speech and language laboratory. 

  Department of Information and Communication Technology
  University of Trento
  Room D11, via Sommarive 14
  38050 Povo di Trento, Italy
  tel : +39-0461 882087
  email: riccardi AT dit.unitn.it

---

Links to Upcoming Conferences and Workshops

(Organized by Date)

IEEE ASRU2005 Automatic Speech Recognition and Understanding Workshop
Cancun, Mexico, November 27 - December 1, 2005
http://www.asru2005.org

ICASSP2006
Toulouse, France May 15-19, 2006
http://www.icassp2006.org

Tutorial and Research Workshop on Speech Recognition and Intrinsic Variation
Toulouse, France May 20, 2006
http://www.divines-project.org/workshop/call.html

2006 Conference on Language Resources and Evaluation
Genoa, Italy May 24-26, 2006
httl://www.lrec-conf.org/lrec2006/

HLT-NAACL 2006
New York City, NY June 4-9, 2006
http://nlp.cs.nyu.edu/hlt-naacl06

Tutorial and Research Workshop on Statistical and Perceptual Audition - SAPA2006
Pittsburgh, PA, USA September 16, 2006
http://www.sapa2006.org/

INTERSPEECH 2006 - ICSLP
Pittsburgh, PA, USA September 17-21, 2006
http://www.interspeech2006.org/

MMSP-2006
Victoria, BC Canada October 3-6, 2006
http://research.microsoft.com/workshops/MMSP06

IEEE/ACL Workshop on SLT
Aruba, Deccember 10-13, 2006
http://www.slt2006.org/CallForPapers.asp

ICASSP2007
Honolulu, Hawaii, USA, 2007, April 17-20

INTERSPEECH 2007
Antwerp, Belgium, August 27-31, 2007
http://www.interspeech2007.org/

back to top

---