Fast Unconstrained Audio Search in Numerous Human Languages

Author

Olsson, J.S. ; Wintrode, Jonathan ; Lee, Minhung

Author_Institution

Dept. of Math., Maryland Univ., College Park, MD, USA

Volume

4

fYear

2007

fDate

15-20 April 2007

Abstract

We present a system to index and search conversational speech using a scoring heuristic on the expected posterior counts of phone n-grams in recognition lattices. We report significant improvements in retrieval effectiveness on five human languages over a strong 1-best baseline. The method is shown to improve the utility (mean average precision) of the retrieved lattices´ rank order and to do so with a search cost negligible compared to the fastest yet known methods for the linear scanning of phonetic lattices.

Keywords

indexing; information retrieval; natural languages; speech processing; speech recognition; mean average precision; numerous human languages; phonetic lattices; recognition lattices; retrieval effectiveness; search conversational speech; unconstrained audio search; Automatic speech recognition; Costs; Hidden Markov models; Humans; Indexing; Information retrieval; Lattices; Natural languages; Speech recognition; Vocabulary; Information retrieval; Natural language interfaces; Natural languages; Speech processing; Speech recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on

Conference_Location

Honolulu, HI

ISSN

1520-6149

Print_ISBN

1-4244-0727-3

Type

conf

DOI

10.1109/ICASSP.2007.367167

Filename

4218041