WNARS：基于WFST的非自回归流端到端语音识别
WNARS: WFST based Non-autoregressive Streaming End-to-End Speech Recognition
摘要：
近年来，基于注意力的编码器-解码器（AED）端到端（E2 E）模型在自动语音识别（ASR）领域引起了越来越多的关注。然而，AED模型在商业应用中部署时仍然具有缺点。自回归束搜索解码使得其对于高并发应用是低效的。集成外部单词级语言模型也不方便。最重要的是，AED模型由于全局注意力机制而难以进行流识别。在本文中，我们提出了一个新的框架，即WNARS，使用混合CTC注意AED模型和加权有限状态传感器（WFST）来解决这些问题。我们从自回归束搜索切换到CTC分支解码，CTC分支解码以块方式使用WFST执行第一遍解码。解码器分支然后对所生成的假设非自回归地执行第二遍重新评分。在AISHELL-1任务上，我们的WNARS实现了5.22%的字符出错率和640 ms的延迟，据我们所知，这是在线ASR的最先进的性能。在我们的10，000小时普通话任务上的进一步实验表明，与强大的TDNN-BLSTM无格MMI基线相比，所提出的方法在50%的延迟下实现了20%以上的改进。
英文摘要：
Recently, attention-based encoder-decoder (AED) end-to-end (E2E) models have drawn more and more attention in the field of automatic speech recognition (ASR). AED models, however, still have drawbacks when deploying in commercial applications. Autoregressive beam search decoding makes it inefficient for high-concurrency applications. It is also inconvenient to integrate external word-level language models. The most important thing is that AED models are difficult for streaming recognition due to global attention mechanism. In this paper, we propose a novel framework, namely WNARS, using hybrid CTC-attention AED models and weighted finite-state transducers (WFST) to solve these problems together. We switch from autoregressive beam search to CTC branch decoding, which performs first-pass decoding with WFST in chunk-wise streaming way. The decoder branch then performs second-pass rescoring on the generated hypotheses non-autoregressively. On the AISHELL-1 task, our WNARS achieves a character error rate of 5.22% with 640ms latency, to the best of our knowledge, which is the state-of-the-art performance for online ASR. Further experiments on our 10,000-hour Mandarin task show the proposed method achieves more than 20% improvements with 50% latency compared to a strong TDNN-BLSTM lattice-free MMI baseline.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
