A new feature extraction front-end for robust speech recognition using progressive histogram equalization and multi-eigenvector temporal filtering.

Shang-nien Tsai, Lin-Shan Lee

Browse the full Interspeech paper archive.