MADI: Inter-domain Matching and Intra-domain Discrimination for Cross-domain Speech Recognition
This addresses domain shift issues in ASR for applications requiring adaptation to new devices or environments, representing a strong incremental improvement over existing methods.
The paper tackles performance degradation in cross-domain speech recognition by proposing MADI, a novel unsupervised domain adaptation approach that combines inter-domain matching and intra-domain discrimination, achieving relative WER reductions of 17.7% and 22.8% on cross-device and cross-environment tasks.
End-to-end automatic speech recognition (ASR) usually suffers from performance degradation when applied to a new domain due to domain shift. Unsupervised domain adaptation (UDA) aims to improve the performance on the unlabeled target domain by transferring knowledge from the source to the target domain. To improve transferability, existing UDA approaches mainly focus on matching the distributions of the source and target domains globally and/or locally, while ignoring the model discriminability. In this paper, we propose a novel UDA approach for ASR via inter-domain MAtching and intra-domain DIscrimination (MADI), which improves the model transferability by fine-grained inter-domain matching and discriminability by intra-domain contrastive discrimination simultaneously. Evaluations on the Libri-Adapt dataset demonstrate the effectiveness of our approach. MADI reduces the relative word error rate (WER) on cross-device and cross-environment ASR by 17.7% and 22.8%, respectively.