VAD-free Streaming Hybrid CTC/Attention ASR for Unsegmented Recording

2021/07/15 by Inaguma, Hirofumi, Kawahara, Tatsuya
#Audio and Speech Processing (eess.AS) #Computation and Language (cs.CL) #FOS: Computer and information sciences #FOS: Electrical engineering #Sound (cs.SD) #electronic engineering #information engineering

paper · doi:10.48550/arxiv.2107.07509

Abstract

In this work, we propose novel decoding algorithms to enable streaming automatic speech recognition (ASR) on unsegmented long-form recordings without voice activity detection (VAD), based on monotonic chunkwise attention (MoChA) with an auxiliary connectionist temporal classification (CTC) objective. We propose a block-synchronous beam search decoding to take advantage of efficient batched output-synchronous and low-latency input-synchronous searches. We also propose a VAD-free inference algorithm that leverages CTC probabilities to determine a suitable timing to reset the model states to tackle the vulnerability to long-form data. Experimental evaluations demonstrate that the block-synchronous decoding achieves comparable accuracy to the label-synchronous one. Moreover, the VAD-free inference can recognize long-form speech robustly for up to a few hours.

Related

• Streaming End-to-End ASR based on Blockwise Non-Autoregressive Models
• End-to-End Automatic Speech Recognition Integrated With CTC-Based Voice Activity Detection
• An Investigation of Enhancing CTC Model for Triggered Attention-based Streaming ASR
• Attention based on-device streaming speech recognition with large speech corpus
• Streaming Audio-Visual Speech Recognition with Alignment Regularization