Speech Retrieval Papers and Project Descriptions

Descriptions of experimental or proprietary systems for which the software may not be available.
AT&T Research
Papers by Julia Hirschberg and others on the SCAN speech-based retrieval and browsing system.
BBN
A brief description of the Rough and Ready speech-based retrieval system.
Cambridge
A description of projects using speech retrieval for video mail and broadcast news by Steve Young and others of the Speech Vision, and Robotics Group in Information Engineering at Cambridge University. Related information and papers from the Olivetti and Oracle Research Lab at Cambridge can be found here. There is a wealth of speech retrieval material on the personal Web pages of the researchers that are reachable from the above links, so it is worth it to spend some time poking around here.
Chinese University of Hong Kong
Work on Spoken Document Retrieval by Helen Meng.
CMU
An extensive set of papers by Alex Hauptmann and others of the Informedia digital library project at Carnegie-Mellon University, several of which describe the use of speech retrieval to select video clips.
Dublin City University
Brief mention of the Taiscéalaí system, which Alan Smeaton completed in 1998.
Johns Hopkins University
Reports from a 6-week workshop on cross-language speech retrieval from the summer of 2000. Johns Hopkins is also a part of the multi-site MALACH project.
LIMSI
Research at the LIMSI spoken language research group in France.
MIT
Kenny Ng's Ph.D. thesis from the MIT Spoken Language Systems Group on speech-based information retrieval using subword indexing. Papers and a Ph.D. dissertation on auditory display for speech retrieval by Barry Arons of the MIT Media Lab are also available.
Michigan State University
The National Gallery of the Spoken Word, a Digital Library Initiative Phase 2 project that plans to apply speech-based retrieval to historical audio collections. There is also a paper on this project available from John Hansen of the University of Colorado.
NIST
A description of the Spoken Interactive Document Retrieval System (SPIDRES) project that was developed at the U.S. National Institute of Standards and Technology. (As prrof that nothing really ever disappears, this page is no longer available at NIST, but it survives from a mirror in Chine.) Several papers on spoken document retrieval are also available from the TREC Web site (see TREC 6-9, and sort by track) and in the DARPA Speech Recognition Workshop and TDT Workshop proceedings, all of which are hosted at NIST.
Royal Melbourne Institute of Technology
Some tutorial information prepared by Corinna Ng.
Ryukoku University
Papers on the retrieval of news video using a combination of speech recognition, speaker identification and face recognition from the lab of Yasuo Ariki and Masahito Kumano at Ryukoku University in Japan (scroll down for English).
University of Maryland
The MALACH project which is developing systems for retrieval from large oral history collections. The project spans 7 organizations, and the main project Web site provides inforamtion about the activites at each location. Some additional papers on Speech Retrieval by Doug Oard are also available.
University of Massachusetts
The keynote address by James Allen from the SIGIR 2001 Workshop on Information Retrieval Techniques for Speech Applications. The full proceedings of that workshop are available online (for a subscription fee) from Springer in Lecture Notes in Computer Science Workshop on Information Retrieval Techniques for Speech Applications.
University of Waterloo
Papers on a system called Jabber that performs speech-based retrieval of recorded meetings by Rick Kazman and others.

Doug Oard
Last modified: Thu Feb 6 00:27:14 2003