Fourth European Conference on Speech Communication and Technology

Madrid, Spain
September 18-21, 1995

Talker-Independent Keyword Spotting for Information Retrieval

J. T. Foote (1), G. J. F. Jones (1,2), K. Sparck Jones (2), S. J. Young (1)

(1) Cambridge University Engineering Department, Cambridge, UK
(2) Cambridge University Computer Laboratory, Cambridge, UK

The goal of the Video Mail Retrieval (VMR) project is to integrate state-of-the-art information retrieval (IR) methods with high-accuracy word spotting to yield a robust and efficient multimedia retrieval system. This paper concerns op en-talker and arbitrary-keyword retrieval based on talker-independent subword models. Because talker-independent subword models can not be expected to work as well as the talker-dependent whole-keyword models used in previous VMR experiments, speaker adaptation is investigated as a means of improving performance (especially for talkers with non-British accents). Both standard FOM word spotting measures and actual retrieval results are computed. The results show that the FOM is not necessarily a good indicator of retrieval performance, and that talker adaptation can substantially improve both spotting and retrieval results.

Full Paper

Bibliographic reference.  Foote, J. T. / Jones, G. J. F. / Jones, K. Sparck / Young, S. J. (1995): "Talker-independent keyword spotting for information retrieval", In EUROSPEECH-1995, 2145-2148.