Publication Details
Title: Longer Features: They Do a Speech Detector Good
Author: TJ Tsai and N. Morgan
Group: Speech
Date: September 2012
PDF: [Not available online]
Overview:
We have incorporated spectrotemporal features in a speech activity detection (SAD) task for the Speech in Noisy Environments 2 (SPINE2) data set. The features were generated by applying 2D Gabor filters to the mel spectrogram in order to measure the strength of various spectral and temporal modulation frequencies in different patches of the spectrogram. Using several different back-ends, the Gabor features significantly outperformed MFCCs, yielding relative reductions in equal error rate (EER) of between 40 and 50%. Compared to the other backends, Adaboost with tree stumps performed particularly well with Gabor features and particularly poorly with MFCCs. An investigation into the reasons for this disparity suggests that the most useful features for SAD incorporate information over longer time scales.
Bibliographic Information:
Proceedings of the 13th Annual Conference of the International Speech Communication Association (InterSpeech 2012), Portland, Oregon
Bibliographic Reference:
TJ Tsai and N. Morgan. Longer Features: They Do a Speech Detector Good. Proceedings of the 13th Annual Conference of the International Speech Communication Association (InterSpeech 2012), Portland, Oregon, September 2012
Author: TJ Tsai and N. Morgan
Group: Speech
Date: September 2012
PDF: [Not available online]
Overview:
We have incorporated spectrotemporal features in a speech activity detection (SAD) task for the Speech in Noisy Environments 2 (SPINE2) data set. The features were generated by applying 2D Gabor filters to the mel spectrogram in order to measure the strength of various spectral and temporal modulation frequencies in different patches of the spectrogram. Using several different back-ends, the Gabor features significantly outperformed MFCCs, yielding relative reductions in equal error rate (EER) of between 40 and 50%. Compared to the other backends, Adaboost with tree stumps performed particularly well with Gabor features and particularly poorly with MFCCs. An investigation into the reasons for this disparity suggests that the most useful features for SAD incorporate information over longer time scales.
Bibliographic Information:
Proceedings of the 13th Annual Conference of the International Speech Communication Association (InterSpeech 2012), Portland, Oregon
Bibliographic Reference:
TJ Tsai and N. Morgan. Longer Features: They Do a Speech Detector Good. Proceedings of the 13th Annual Conference of the International Speech Communication Association (InterSpeech 2012), Portland, Oregon, September 2012