Towards Effective and Efficient Non-autoregressive Decoding Using Block-based Attention Mask

• URL: http://arxiv.org/abs/2406.10034v2
• Date: Tue, 16 Jul 2024 06:34:46 GMT
• Title: Towards Effective and Efficient Non-autoregressive Decoding Using Block-based Attention Mask
• Authors: Tianzi Wang, Xurong Xie, Zhaoqing Li, Shoukang Hu, Zengrui Jing, Jiajun Deng, Mingyu Cui, Shujie Hu, Mengzhe Geng, Guinan Li, Helen Meng, Xunying Liu,
• Abstract summary: AMD performs parallel NAR inference within contiguous blocks of output labels concealed using attention masks. A beam search algorithm is designed to leverage a dynamic fusion of CTC, AR Decoder, and AMD probabilities. Experiments on the LibriSpeech-100hr corpus suggest the tripartite Decoder incorporating the AMD module produces a maximum decoding speed-up ratio of 1.73x.
• Score: 73.5088336210358
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: This paper proposes a novel non-autoregressive (NAR) block-based Attention Mask Decoder (AMD) that flexibly balances performance-efficiency trade-offs for Conformer ASR systems. AMD performs parallel NAR inference within contiguous blocks of output labels that are concealed using attention masks, while conducting left-to-right AR prediction and history context amalgamation between blocks. A beam search algorithm is designed to leverage a dynamic fusion of CTC, AR Decoder, and AMD probabilities. Experiments on the LibriSpeech-100hr corpus suggest the tripartite Decoder incorporating the AMD module produces a maximum decoding speed-up ratio of 1.73x over the baseline CTC+AR decoding, while incurring no statistically significant word error rate (WER) increase on the test sets. When operating with the same decoding real time factors, statistically significant WER reductions of up to 0.7% and 0.3% absolute (5.3% and 6.1% relative) were obtained over the CTC+AR baseline.

Related papers

• Non-autoregressive End-to-end Speech Translation with Parallel Autoregressive Rescoring [83.32560748324667]
  This article describes an efficient end-to-end speech translation (E2E-ST) framework based on non-autoregressive (NAR) models. We propose a unified NAR E2E-ST framework called Orthros, which has an NAR decoder and an auxiliary shallow AR decoder on top of the shared encoder.
  arXiv  Detail & Related papers  (2021-09-09T16:50:16Z)
• Streaming End-to-End ASR based on Blockwise Non-Autoregressive Models [57.20432226304683]
  Non-autoregressive (NAR) modeling has gained more and more attention in speech processing. We propose a novel end-to-end streaming NAR speech recognition system. We show that the proposed method improves online ASR recognition in low latency conditions.
  arXiv  Detail & Related papers  (2021-07-20T11:42:26Z)
• Non-Autoregressive Transformer ASR with CTC-Enhanced Decoder Input [54.82369261350497]
  We propose a CTC-enhanced NAR transformer, which generates target sequence by refining predictions of the CTC module. Experimental results show that our method outperforms all previous NAR counterparts and achieves 50x faster decoding speed than a strong AR baseline with only 0.0 0.3 absolute CER degradation on Aishell-1 and Aishell-2 datasets.
  arXiv  Detail & Related papers  (2020-10-28T15:00:09Z)
• Improved Mask-CTC for Non-Autoregressive End-to-End ASR [49.192579824582694]
  Recently proposed end-to-end ASR system based on mask-predict with connectionist temporal classification (CTC) We propose to enhance the network architecture by employing a recently proposed architecture called Conformer. Next, we propose new training and decoding methods by introducing auxiliary objective to predict the length of a partial target sequence.
  arXiv  Detail & Related papers  (2020-10-26T01:22:35Z)
• FastLR: Non-Autoregressive Lipreading Model with Integrate-and-Fire [74.04394069262108]
  We propose FastLR, a non-autoregressive (NAR) lipreading model which generates all target tokens simultaneously. FastLR achieves the speedup up to 10.97$times$ compared with state-of-the-art lipreading model.
  arXiv  Detail & Related papers  (2020-08-06T08:28:56Z)
• Fault-Tolerant Weighted Union-Find Decoding on the Toric Code [2.492300648514129]
  We benchmark a weighted variant of the union-find decoder on the toric code under circuit-level depolarizing noise. This variant preserves the almost-linear time complexity of the original while significantly increasing the performance in the fault-tolerance setting.
  arXiv  Detail & Related papers  (2020-04-09T17:24:30Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.