Robust features derived from temporal trajectory filtering for speech recognition under the corruption of additive and convolutional noises.

Kuo-Hwei Yuo, Hsiao-Chuan Wang

Browse the full ICASSP paper archive.