NAR-MBR decoding applies minimum Bayes risk selection to samples drawn efficiently from a non-autoregressive ASR model, yielding higher accuracy than prior NAR methods while remaining faster than autoregressive decoding.
Non-Autoregressive Minimum Bayes' Risk Decoding for Fast Speech Recognition
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
Non-autoregressive (NAR) decoding generates output tokens in parallel, making speech recognition faster than autoregressive decoding, which generates them sequentially from left to right. However, the recognition performance is degraded because NAR decoding cannot resolve uncertainty by conditioning on previously generated tokens. To address this issue, we propose a novel NAR decoding framework based on minimum Bayes' risk (MBR) decoding, termed NAR-MBR decoding, that maximizes the expected utility calculated from samples drawn from the output probability of an NAR model rather than maximizing the output probability. Notably, by leveraging the nature of NAR models, multiple samples are obtained efficiently with a single forward computation. Our experiments across LibriSpeech, Switchboard, AMI, and web presentation corpus demonstrated that our NAR-MBR decoding outperformed previous NAR decoding and ran faster than AR decoding.
fields
eess.AS 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Non-Autoregressive Minimum Bayes' Risk Decoding for Fast Speech Recognition
NAR-MBR decoding applies minimum Bayes risk selection to samples drawn efficiently from a non-autoregressive ASR model, yielding higher accuracy than prior NAR methods while remaining faster than autoregressive decoding.