Matches in SemOpenAlex for { <https://semopenalex.org/work/W3184187848> ?p ?o ?g. }
- W3184187848 abstract "Non-autoregressive (NAR) modeling has gained more and more attention in speech processing. With recent state-of-the-art attention-based automatic speech recognition (ASR) structure, NAR can realize promising real-time factor (RTF) improvement with only small degradation of accuracy compared to the autoregressive (AR) models. However, the recognition inference needs to wait for the completion of a full speech utterance, which limits their applications on low latency scenarios. To address this issue, we propose a novel end-to-end streaming NAR speech recognition system by combining blockwise-attention and connectionist temporal classification with mask-predict (Mask-CTC) NAR. During inference, the input audio is separated into small blocks and then processed in a blockwise streaming way. To address the insertion and deletion error at the edge of the output of each block, we apply an overlapping decoding strategy with a dynamic mapping trick that can produce more coherent sentences. Experimental results show that the proposed method improves online ASR recognition in low latency conditions compared to vanilla Mask-CTC. Moreover, it can achieve a much faster inference speed compared to the AR attention-based models. All of our codes will be publicly available at https://github.com/espnet/espnet." @default.
- W3184187848 created "2021-08-02" @default.
- W3184187848 creator A5001291873 @default.
- W3184187848 creator A5040732498 @default.
- W3184187848 creator A5050058892 @default.
- W3184187848 creator A5088539115 @default.
- W3184187848 date "2021-07-20" @default.
- W3184187848 modified "2023-09-23" @default.
- W3184187848 title "Streaming End-to-End ASR based on Blockwise Non-Autoregressive Models" @default.
- W3184187848 cites W1524333225 @default.
- W3184187848 cites W1586532344 @default.
- W3184187848 cites W1828163288 @default.
- W3184187848 cites W2127141656 @default.
- W3184187848 cites W2739427748 @default.
- W3184187848 cites W2767206889 @default.
- W3184187848 cites W2773781902 @default.
- W3184187848 cites W2911109671 @default.
- W3184187848 cites W2936774411 @default.
- W3184187848 cites W2951655021 @default.
- W3184187848 cites W2962780374 @default.
- W3184187848 cites W2963242190 @default.
- W3184187848 cites W2963827914 @default.
- W3184187848 cites W2972818416 @default.
- W3184187848 cites W2987019345 @default.
- W3184187848 cites W2989134874 @default.
- W3184187848 cites W3008898571 @default.
- W3184187848 cites W3014413043 @default.
- W3184187848 cites W3015927303 @default.
- W3184187848 cites W3015974384 @default.
- W3184187848 cites W3024732798 @default.
- W3184187848 cites W3025165719 @default.
- W3184187848 cites W3025417467 @default.
- W3184187848 cites W3026287411 @default.
- W3184187848 cites W3035445001 @default.
- W3184187848 cites W3092122846 @default.
- W3184187848 cites W3093700231 @default.
- W3184187848 cites W3094800360 @default.
- W3184187848 cites W3097224945 @default.
- W3184187848 cites W3148654612 @default.
- W3184187848 cites W3149509723 @default.
- W3184187848 cites W3162431424 @default.
- W3184187848 cites W3162899666 @default.
- W3184187848 doi "https://doi.org/10.48550/arxiv.2107.09428" @default.
- W3184187848 hasPublicationYear "2021" @default.
- W3184187848 type Work @default.
- W3184187848 sameAs 3184187848 @default.
- W3184187848 citedByCount "0" @default.
- W3184187848 crossrefType "posted-content" @default.
- W3184187848 hasAuthorship W3184187848A5001291873 @default.
- W3184187848 hasAuthorship W3184187848A5040732498 @default.
- W3184187848 hasAuthorship W3184187848A5050058892 @default.
- W3184187848 hasAuthorship W3184187848A5088539115 @default.
- W3184187848 hasBestOaLocation W31841878481 @default.
- W3184187848 hasConcept C11413529 @default.
- W3184187848 hasConcept C137293760 @default.
- W3184187848 hasConcept C149782125 @default.
- W3184187848 hasConcept C154945302 @default.
- W3184187848 hasConcept C159877910 @default.
- W3184187848 hasConcept C162324750 @default.
- W3184187848 hasConcept C2775852435 @default.
- W3184187848 hasConcept C2776214188 @default.
- W3184187848 hasConcept C28490314 @default.
- W3184187848 hasConcept C41008148 @default.
- W3184187848 hasConcept C50644808 @default.
- W3184187848 hasConcept C57273362 @default.
- W3184187848 hasConcept C74296488 @default.
- W3184187848 hasConcept C76155785 @default.
- W3184187848 hasConcept C82876162 @default.
- W3184187848 hasConcept C8521452 @default.
- W3184187848 hasConceptScore W3184187848C11413529 @default.
- W3184187848 hasConceptScore W3184187848C137293760 @default.
- W3184187848 hasConceptScore W3184187848C149782125 @default.
- W3184187848 hasConceptScore W3184187848C154945302 @default.
- W3184187848 hasConceptScore W3184187848C159877910 @default.
- W3184187848 hasConceptScore W3184187848C162324750 @default.
- W3184187848 hasConceptScore W3184187848C2775852435 @default.
- W3184187848 hasConceptScore W3184187848C2776214188 @default.
- W3184187848 hasConceptScore W3184187848C28490314 @default.
- W3184187848 hasConceptScore W3184187848C41008148 @default.
- W3184187848 hasConceptScore W3184187848C50644808 @default.
- W3184187848 hasConceptScore W3184187848C57273362 @default.
- W3184187848 hasConceptScore W3184187848C74296488 @default.
- W3184187848 hasConceptScore W3184187848C76155785 @default.
- W3184187848 hasConceptScore W3184187848C82876162 @default.
- W3184187848 hasConceptScore W3184187848C8521452 @default.
- W3184187848 hasLocation W31841878481 @default.
- W3184187848 hasOpenAccess W3184187848 @default.
- W3184187848 hasPrimaryLocation W31841878481 @default.
- W3184187848 hasRelatedWork W2937649809 @default.
- W3184187848 hasRelatedWork W2996122240 @default.
- W3184187848 hasRelatedWork W3141854550 @default.
- W3184187848 hasRelatedWork W3156915121 @default.
- W3184187848 hasRelatedWork W3184187848 @default.
- W3184187848 hasRelatedWork W3197304116 @default.
- W3184187848 hasRelatedWork W3199016780 @default.
- W3184187848 hasRelatedWork W4285757700 @default.
- W3184187848 hasRelatedWork W4319862213 @default.
- W3184187848 hasRelatedWork W4366851091 @default.
- W3184187848 isParatext "false" @default.
- W3184187848 isRetracted "false" @default.