Noise robust keyword spotting for user generated video blogs

This paper presents a template-based system for speaker independent key word spotting (KWS) in continuous speech that can help in automatic analysis, indexing, search and retrieval of user generated videos by content. Extensive experiments on clean speech confirm that the proposed approach is superi...

Full description

Saved in:
Bibliographic Details
Main Authors: Barakat, M. S., Ritz, C. H., Stirling, D. A.
Format: Conference Proceeding
Language:English
Subjects:
Blogs
Hidden Markov models
Histograms
Keyword Spotting (KWS)
Noise measurement
Noise Robustness
Signal to noise ratio
Social Networks
Speech
Template Matching (TM)
Users Video Blogs
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper presents a template-based system for speaker independent key word spotting (KWS) in continuous speech that can help in automatic analysis, indexing, search and retrieval of user generated videos by content. Extensive experiments on clean speech confirm that the proposed approach is superior to a HMM approach when applied to noisy speech with different signal-to-noise ratio (SNR) levels. Experiments conducted to detect swear words, personal names and product names within a set of online user generated video blogs shows significantly better recall and precision results compared to a traditional ASR-based approach.
ISSN:1945-7871
1945-788X
DOI:10.1109/ICME.2013.6607589