Sub-word modeling of out of vocabulary words in spoken term detection

This paper deals with comparison of sub-word based methods for spoken term detection (STD) task and phone recognition. The sub-word units are needed for search for out-of-vocabulary words. We compared words, phones and multigrams. The maximal length and pruning of multigrams were investigated first....

Full description

Saved in:
Bibliographic Details
Main Authors: Szoke, I., Burget, L., Cernocky, J., Fapso, M.
Format: Conference Proceeding
Language:eng ; jpn
Subjects:
Broadcasting
Dictionaries
Indexing
Information technology
keyword spotting
lattice
Lattices
multigram
NIST
phone
Speech processing
Speech recognition
spoken term detection
sub-word
syllable
Telephony
Vocabulary
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper deals with comparison of sub-word based methods for spoken term detection (STD) task and phone recognition. The sub-word units are needed for search for out-of-vocabulary words. We compared words, phones and multigrams. The maximal length and pruning of multigrams were investigated first. Then two constrained methods of multigram training were proposed. We evaluated on the NIST STD06 dev-set CTS data. The conclusion is that the proposed method improves the phone accuracy more than 9% relative and STD accuracy more than 7% relative.
DOI:10.1109/SLT.2008.4777893