Multiple Acoustic Model-Based Discriminative Likelihood Ratio Weighting for Voice Activity Detection

2012 ◽  
Vol 19 (8) ◽  
pp. 507-510 ◽  
Author(s):  
Youngjoo Suh ◽  
Hoirin Kim
Keyword(s):  
Likelihood Ratio ◽  
Voice Activity Detection ◽  
Acoustic Model ◽  
Activity Detection ◽  
Model Based ◽  
Voice Activity

A statistical model-based voice activity detection employing minimum classification error technique

2008 ◽  
Author(s):  
Sang-Ick Kang ◽  
Ji-Hyun Song ◽  
Kye-Hwan Lee ◽  
Yun-Sik Park ◽  
Joon-Hyuk Chang
Keyword(s):  
Statistical Model ◽  
Voice Activity Detection ◽  
Classification Error ◽  
Activity Detection ◽  
Model Based ◽  
Minimum Classification Error ◽  
Voice Activity

Noise robust model-based voice activity detection

2006 ◽  
Author(s):  
Ángel de la Torre ◽  
Javier Ramírez ◽  
Carmen Benítez ◽  
José C. Segura ◽  
L. García ◽  
...  
Keyword(s):  
Voice Activity Detection ◽  
Activity Detection ◽  
Model Based ◽  
Robust Model ◽  
Noise Robust ◽  
Voice Activity

Discriminative Weight Training for a Statistical Model-Based Voice Activity Detection

2008 ◽  
Vol 15 ◽  
pp. 170-173 ◽  
Author(s):  
Sang Ick Kang ◽  
Q.H. Jo ◽  
Joon Hyuk Chang
Keyword(s):  
Statistical Model ◽  
Weight Training ◽  
Voice Activity Detection ◽  
Activity Detection ◽  
Model Based ◽  
Voice Activity

Likelihood ratio sign test for voice activity detection

2012 ◽  
Vol 6 (4) ◽  
pp. 306 ◽  
Author(s):  
S. Deng ◽  
J. Han
Keyword(s):  
Likelihood Ratio ◽  
Voice Activity Detection ◽  
Sign Test ◽  
Activity Detection ◽  
Voice Activity

scholarly journals Combining augmented statistical noise suppression and framewise speech/non-speech classification for robust voice activity detection

2017 ◽  
Vol 6 ◽  
Author(s):  
Yasunari Obuchi
Keyword(s):  
Noise Suppression ◽  
Frequency Component ◽  
Voice Activity Detection ◽  
Training Data ◽  
Activity Detection ◽  
Noisy Environments ◽  
Model Based ◽  
Speech Classification ◽  
Voice Activity ◽  
Statistical Noise

This paper proposes a new voice activity detection (VAD) algorithm based on statistical noise suppression and framewise speech/non-speech classification. Although many VAD algorithms have been developed that are robust in noisy environments, the most successful ones are related to statistical noise suppression in some way. Accordingly, we formulate our VAD algorithm as a combination of noise suppression and subsequent framewise classification. The noise suppression part is improved by introducing the idea that any unreliable frequency component should be removed, and the decision can be made by the remaining signal. This augmentation can be realized using a few additional parameters embedded in the gain-estimation process. The framewise classification part can be either model-less or model-based. A model-less classifier has the advantage that it can be applied to any situation, even if no training data are available. In contrast, a model-based classifier (e.g., neural network-based classifier) requires training data but tends to be more accurate. The accuracy of the proposed algorithm is evaluated using the CENSREC-1-C public framework and confirmed to be superior to many existing algorithms.

Sign in / Sign up

Export Citation Format

Share Document