A Front-End Speech Enhancement System for Robust Automotive Speech Recognition
Haikun Wang, Zhongfu Ye, Jingdong Chen · 2018
This paper presents a front-end speech enhancement approach to robust speech recognition in automotive environments. It combines model-based voice activity detection (VAD), relative transfer function (RTF) based generalized sidelobe cancelation, and single-channel post filtering to enhance the speech signal of interest, thereby improving the robustness of speech recognition. First, we choose four typical driving scenarios, which include most of the noise types in automobiles to record training data. The recorded data are then used to train Gaussian mixture models (GMMs) for both speech and noise. The trained GMMs are subsequently used to estimate the speech presence probability on a frame-by-frame basis. This speech presence probability is then served as the basic information for RTF estimation, adaptive beamforming, and post-filtering. Experiments are conducted in real automotive environments and the results show that the developed method can significantly improve the performance of both VAD and automatic speech recognition (ASR).