Uncertainty driven Compensation of Multi-Stream MLP Acoustic Models for Robust ASR

摘要

In this paper we show how the robustness of multi-stream multi-layer perceptron (MLP) acoustic models can be increased through uncertainty propagation and decoding. We demonstrate that MLP uncertainty decoding yields consistent improvements over using minimum mean square error (MMSE) feature enhancement in MFCC and RASTA-LPCC domains. We introduce as well formulas for the computation of the uncertainty associated to the acoustic likelihood computation and explore different stream integration schemes using this uncertainty on the AURORA4 corpus.