1 |
Error Correction in ASR using Sequence-to-Sequence Models ...
|
|
|
|
Abstract:
Post-editing in Automatic Speech Recognition (ASR) entails automatically correcting common and systematic errors produced by the ASR system. The outputs of an ASR system are largely prone to phonetic and spelling errors. In this paper, we propose to use a powerful pre-trained sequence-to-sequence model, BART, further adaptively trained to serve as a denoising model, to correct errors of such types. The adaptive training is performed on an augmented dataset obtained by synthetically inducing errors as well as by incorporating actual errors from an existing ASR system. We also propose a simple approach to rescore the outputs using word level alignments. Experimental results on accented speech data demonstrate that our strategy effectively rectifies a significant number of ASR errors and produces improved WER results when compared against a competitive baseline. ...
|
|
Keyword:
Computation and Language cs.CL; FOS Computer and information sciences; Machine Learning cs.LG
|
|
URL: https://dx.doi.org/10.48550/arxiv.2202.01157 https://arxiv.org/abs/2202.01157
|
|
BASE
|
|
Hide details
|
|
2 |
Accurate Online Posterior Alignments for Principled Lexically-Constrained Decoding ...
|
|
|
|
BASE
|
|
Show details
|
|
3 |
From Machine Translation to Code-Switching: Generating High-Quality Code-Switched Text ...
|
|
|
|
BASE
|
|
Show details
|
|
4 |
Rudder: A Cross Lingual Video and Text Retrieval Dataset ...
|
|
|
|
BASE
|
|
Show details
|
|
5 |
Meta-Learning for Effective Multi-task and Multilingual Modelling ...
|
|
|
|
BASE
|
|
Show details
|
|
6 |
The Effectiveness of Intermediate-Task Training for Code-Switched Natural Language Understanding ...
|
|
|
|
BASE
|
|
Show details
|
|
7 |
Multilingual and code-switching ASR challenges for low resource Indian languages ...
|
|
|
|
BASE
|
|
Show details
|
|
8 |
From Machine Translation to Code-Switching: Generating High-Quality Code-Switched Text ...
|
|
|
|
BASE
|
|
Show details
|
|
9 |
The Effectiveness of Intermediate-Task Training for Code-Switched Natural Language Understanding ...
|
|
|
|
BASE
|
|
Show details
|
|
10 |
Automatic Speech Recognition in Sanskrit: A New Speech Corpus and Modelling Insights ...
|
|
|
|
BASE
|
|
Show details
|
|
11 |
Automatic Speech Recognition in Sanskrit: A New Speech Corpus and Modelling Insights ...
|
|
|
|
BASE
|
|
Show details
|
|
12 |
Improving Low Resource Code-switched ASR using Augmented Code-switched TTS ...
|
|
|
|
BASE
|
|
Show details
|
|
13 |
Reduce and Reconstruct: ASR for Low-Resource Phonetic Languages ...
|
|
|
|
BASE
|
|
Show details
|
|
14 |
Dual Language Models for Code Switched Speech Recognition ...
|
|
|
|
BASE
|
|
Show details
|
|
|
|