paper-with-me

홈 › Papers

Non-autoregressive Error Correction for CTC-based ASR with Phone-conditioned Masked LM

2022-09-08 · Hayato Futami, Hirofumi Inaguma, Sei Ueno, Masato Mimura, Shinsuke Sakai, Tatsuya Kawahara

Connectionist temporal classification (CTC) -based models are attractive in automatic speech recognition (ASR) because of their non-autoregressive nature. To take advantage of text-only data, language model (LM) integration approaches such as rescoring and shallow fusion have been widely used for CTC. However, they lose CTC's non-autoregressive nature because of the need for beam search, which slows down the inference speed. In this study, we propose an error correction method with phone-conditioned masked LM (PC-MLM). In the proposed method, less confident word tokens in a greedy decoded output from CTC are masked. PC-MLM then predicts these masked word tokens given unmasked words and phones supplementally predicted from CTC. We further extend it to Deletable PC-MLM in order to address insertion errors. Since both CTC and PC-MLM are non-autoregressive models, the method enables fast LM integration. Experimental evaluations on the Corpus of Spontaneous Japanese (CSJ) and TED-LIUM2 in domain adaptation setting shows that our proposed method outperformed rescoring and shallow fusion in terms of inference speed, and also in terms of recognition accuracy on CSJ.

📄 PDF Abstract BibTeX arXiv:2209.04062

Code (1)

emonosuke/emoasr 공식 구현 pytorch

Tasks

Automatic Speech RecognitionAutomatic Speech Recognition (ASR)Domain AdaptationLanguage ModelingLanguage Modellingspeech-recognitionSpeech Recognition

Similar Papers 제목 키워드 기반

Graph-Based Phonetic Error Correction of Noisy ASR

2026-04-29 · Pratik Rakesh Singh, Mohammadi Zaki, Aneesh Mukkamala, Pankaj Wasnik arxiv

Automatic speech recognition (ASR) systems, despite low overall word error rates, produce residual lexical errors that disproportionately affect semantically critical tokens such as named entities, negations, and sentime…

Graph Neural NetworkSpeech Recognition

Non-Autoregressive Chinese ASR Error Correction with Phonological Training

2022-07-01 · NAACL 2022 7 · Zheng Fang, Ruiqing Zhang, Zhongjun He, Hua Wu 외

Automatic Speech Recognition (ASR) is an efficient and widely used input method that transcribes speech signals into text. As the errors introduced by ASR systems will impair the performance of downstream tasks, we intro…

Automatic Speech RecognitionAutomatic Speech Recognition (ASR)Sentencespeech-recognition+1

Integrated Semantic and Phonetic Post-correction for Chinese Speech Recognition

2021-11-16 · ROCLING 2021 10 · Yi-Chang Chen, Chun-Yen Cheng, Chien-An Chen, Ming-Chieh Sung 외

Due to the recent advances of natural language processing, several works have applied the pre-trained masked language model (MLM) of BERT to the post-correction of speech recognition. However, existing pre-trained models…

Language ModelingLanguage Modellingspeech-recognitionSpeech Recognition

MaskGCT: Zero-Shot Text-to-Speech with Masked Generative Codec Transformer

2024-09-01 · Yuancheng Wang, Haoyue Zhan, Liwei Liu, Ruihong Zeng 외

The recent large-scale text-to-speech (TTS) systems are usually grouped as autoregressive and non-autoregressive systems. The autoregressive systems implicitly model duration but exhibit certain deficiencies in robustnes…

Self-Supervised Learningtext-to-speechText to Speech

DCGC: Draft-Conditioned Global Correction for Complex Reasoning with Masked Diffusion Models

2026-08-26 · Minhae Oh, Nakyung Lee, Jungwoo Lee arxiv

Correcting flawed reasoning traces remains a significant challenge for Large Language Models (LLMs), whose autoregressive generation can propagate early mistakes into subsequent reasoning. We introduce DCGC, a Masked Dif…