A novel feature-extraction for speech recognition based on multiple acoustic-feature planes.

Tsuneo Nitta

Browse the full ICASSP paper archive.