Re-evaluating Minimum Bayes Risk Decoding for Automatic Speech Recognition
This work addresses the need for more accurate decoding in speech-to-text tasks like ASR and ST, though it is incremental as it applies an existing method to new domains.
The paper tackled the problem of improving decoding methods for automatic speech recognition and speech translation by evaluating Minimum Bayes Risk (MBR) decoding against beam search, finding that MBR outperforms beam search in most settings for higher accuracy in offline tasks.
Recent work has shown that sample-based Minimum Bayes Risk (MBR) decoding outperforms beam search in text-to-text generation tasks, such as machine translation, text summarization, and image captioning. On the other hand, beam search is the current practice for speech-to-text tasks such as automatic speech recognition (ASR) and Speech Translation (ST). Given that MBR decoding is effective in text-to-text generation tasks, it is reasonable to expect it to also be effective for speech-to-text tasks. In this paper, we evaluate MBR decoding for ASR and ST tasks on English and Japanese using Whisper and its derivative models. We observe that the accuracy of MBR decoding outperforms that of beam search in most of the experimental settings we have evaluated. The results show that MBR decoding is a promising method for offline ASR and ST tasks that require high accuracy. The code is available at https://github.com/CyberAgentAILab/mbr-for-asr