Robust speech endpoint detection based on improved adaptive band-partitioning spectral entropy

Authors:
Xin Li;Huaping Liu;Yu Zheng;Bolin Xu
Affiliations:
Dept. of Elect. and Inf. Eng., Nanjing Univ., Nanjing, China and State Key Lab of Pattern Recognition, Inst. of Automation, Chinese Academy of Sci., Beijing, China and Sch. of Electromech. Eng. an ...;School of Electromechanical Engineering and Automation, Shanghai University, Shanghai, China;School of Computer Engineering and Science, Shanghai University, Shanghai, China;Department of Electronic and Information Engineering, Nanjing University, Nanjing, China
Venue:
LSMS'07 Proceedings of the Life system modeling and simulation 2007 international conference on Bio-Inspired computational intelligence and applications
Year:
2007

Citing 1
Cited 2

A robust algorithm for accurate endpointing of speech signals

Speech Communication

Statistical analysis of complementary spectral features of emotional speech in Czech and Slovak

TSD'11 Proceedings of the 14th international conference on Text, speech and dialogue
Comparison of complementary spectral features of emotional speech for german, czech, and slovak

COST'11 Proceedings of the 2011 international conference on Cognitive Behavioural Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

The performance of speech recognition system is often degraded in adverse environments. Accurate Speech endpoint detection is very important for robust speech recognition. In this paper, an improved adaptive band-partitioning spectral entropy algorithm was proposed for speech endpoint detection, which utilized the weighted power spectral subtraction to boost up the signal-to-noise ratio (SNR) as well as keep the robustness. The idea of adaptive band-partitioning spectral entropy is to divide a frame into some sub-bands which the number of it could be selected adaptively, and calculate spectral entropy of them. Although it has good robustness, the accuracy degrades rapidly when the SNR are low. Therefore, the weighted power spectral subtraction is presented for reducing the spectral effects of acoustically added noise in speech. The speech recognition experiment results indicate that the recognition accuracy have improved well in adverse environments.