Cross-domain Speech Recognition with Unsupervised Character-level Distribution Matching

15 Apr 2021  ·  Wenxin Hou, Jindong Wang, Xu Tan, Tao Qin, Takahiro Shinozaki ·

End-to-end automatic speech recognition (ASR) can achieve promising performance with large-scale training data. However, it is known that domain mismatch between training and testing data often leads to a degradation of recognition accuracy... In this work, we focus on the unsupervised domain adaptation for ASR and propose CMatch, a Character-level distribution matching method to perform fine-grained adaptation between each character in two domains. First, to obtain labels for the features belonging to each character, we achieve frame-level label assignment using the Connectionist Temporal Classification (CTC) pseudo labels. Then, we match the character-level distributions using Maximum Mean Discrepancy. We train our algorithm using the self-training technique. Experiments on the Libri-Adapt dataset show that our proposed approach achieves 14.39% and 16.50% relative Word Error Rate (WER) reduction on both cross-device and cross-environment ASR. We also comprehensively analyze the different strategies for frame-level label assignment and Transformer adaptations. read more

PDF Abstract

Datasets


Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Cross-environment ASR Libri-Adapt CMatch Word Error Rate (WER) 28.19 # 1
Cross-device ASR Libri-Adapt CMatch Word Error Rate (WER) 22.85 # 1

Methods