Paper Info
Title
The Kaldi Openkws System: Improving Low Resource Keyword Search
Abstract
The IARPA BABEL program has stimulated worldwide research in keyword search technology for low resource languages, and the NIST OpenKWS evaluations are the de facto benchmark test for such capabilities. The 2016 OpenKWS evaluation featured Georgian speech, and had 10 participants from across the world. This paper describes the Kaldi system developed to assist IARPA in creating a competitive baseline against which participants were evaluated, and to provide a truly open source system to all participants to support their research. This system handily met the BABEL program goals of 0.60 ATWV and 50% WER, achieving 0.70 ATWV and 38% WER with a single ASR system, i.e. without ASR system combination. All except one OpenKWS participant used Kaldi components in their submissions, typically in conjunction with system combination. This paper therefore complements all other OpenKWS-based papers.
Year
DOI
Venue
2017
10.21437/Interspeech.2017-601
18TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2017), VOLS 1-6: SITUATED INTERACTION
Keywords
Field
DocType
speech recognition, keyword search, spoken term detection, IARPA Babel, OpenKWS
Computer science,Keyword search,Speech recognition
Conference
ISSN
Citations 
PageRank 
2308-457X
3
0.41
References 
Authors
12
10
Name
Order
Citations
PageRank
Jan Trmal123520.91
Matthew Wiesner252.85
Vijayaditya Peddinti322912.17
Xiaohui Zhang419419.81
Pegah Ghahremani5997.09
Yiming Wang6173.27
Vimal Manohar7547.99
Hainan Xu8145.56
Daniel Povey92442231.75
Sanjeev Khudanpur102155202.00