Spatial Diffuseness Features for DNN-Based Speech Recognition in Noisy and Reverberant Environments

Abstract

We propose a spatial diffuseness feature for deep neural network (DNN)-based automatic speech recognition to improve recognition accuracy in reverberant and noisy environments. The feature is computed in real-time from multiple microphone signals without requiring knowledge or estimation of the direction of arrival, and represents the relative amount of diffuse noise in each time and frequency bin. It is shown that using the diffuseness feature as an additional input to a DNN-based acoustic model leads to a reduced word error rate for the REVERB challenge corpus, both compared to logmelspec features extracted from noisy signals, and features enhanced by spectral subtraction.

accepted for ICASSP2015

References in corpus (1)

• Coherent-to-Diffuse Power Ratio Estimation for Dereverberation

Cited by in corpus (6)

• Light Gated Recurrent Units for Speech Recognition
• Robust sound event detection in bioacoustic sensor networks
• Coherent-to-Diffuse Power Ratio Estimation for Dereverberation
• Robust coherence-based spectral enhancement for speech recognition in adverse real-world environments
• Robust coherence-based spectral enhancement for distant speech recognition
• An improved uncertainty decoding scheme with weighted samples for DNN-HMM hybrid systems