Statistical Error Correction Methods for Domain-Specific ASR Systems

Cucu, Horia; Buzo, Andi; Besacier, Laurent; Burileanu, Corneliu

doi:10.1007/978-3-642-39593-2_7

Cited by 18 publications

(15 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, each component must be implemented independently (Paulik et al, 2008;Škodová et al, 2012). Some post-processing studies have been conducted using the N-gram language model; however, the statistics-based method requires a large amount data and cannot consider the context (Cucu et al, 2013;Bassil and Semaan, 2012).…”

Section: Conventional Methodologymentioning

confidence: 99%

BTS: Back TranScription for Speech-to-Text Post-Processor using Text-to-Speech-to-Text

Park¹,

Seo²,

Lee³

et al. 2021

Proceedings of the 8th Workshop on Asian Translation (WAT2021)

View full text Add to dashboard Cite

With the growing popularity of smart speakers, such as Amazon Alexa, speech is becoming one of the most important modes of humancomputer interaction. Automatic speech recognition (ASR) is arguably the most critical component of such systems, as errors in speech recognition propagate to the downstream components and drastically degrade the user experience. A simple and effective way to improve the speech recognition accuracy is to apply automatic post-processor to the recognition result. However, training a post-processor requires parallel corpora created by human annotators, which are expensive and not scalable. To alleviate this problem, we propose Back TranScription (BTS), a denoising-based method that can create such corpora without human labor. Using a raw corpus, BTS corrupts the text using Text-to-Speech (TTS) and Speech-to-Text (STT) systems. Then, a postprocessing model can be trained to reconstruct the original text given the corrupted input. Quantitative and qualitative evaluations show that a post-processor trained using our approach is highly effective in fixing non-trivial speech recognition errors such as mishandling foreign words. We present the generated parallel corpus and post-processing platform to make our results publicly available.

show abstract

Section: Conventional Methodologymentioning

confidence: 99%

BTS: Back TranScription for Speech-to-Text Post-Processor using Text-to-Speech-to-Text

Park¹,

Seo²,

Lee³

et al. 2021

Proceedings of the 8th Workshop on Asian Translation (WAT2021)

View full text Add to dashboard Cite

show abstract

“…Early work [12] proposed using domainspecific pre-parsed "exemplar sentences", which might be enough for closed domains with a small number of possible intents. Another approach is using a statistical machine translation model trained on raw/corrected ASR output pairs to correct future errors [3]. Recent work uses language models to correct the output of ASR systems.…”

Section: Related Workmentioning

confidence: 99%

Mondegreen: A Post-Processing Solution to Speech Recognition Error Correction for Voice Search Queries

Sodhi

Chio

Jash

et al. 2021

Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery &Amp; Data Mining

View full text Add to dashboard Cite

As more and more online search queries come from voice, automatic speech recognition becomes a key component to deliver relevant search results. Errors introduced by automatic speech recognition (ASR) lead to irrelevant search results returned to the user, thus causing user dissatisfaction. In this paper, we introduce an approach, Mondegreen, to correct voice queries in text space without depending on audio signals, which may not always be available due to system constraints or privacy or bandwidth (for example, some ASR systems run on-device) considerations. We focus on voice queries transcribed via several proprietary commercial ASR systems. These queries come from users making internet, or online service search queries. We first present an analysis showing how different the language distribution coming from user voice queries is from that in traditional text corpora used to train off-the-shelf ASR systems. We then demonstrate that Mondegreen can achieve significant improvements in increased user interaction by correcting user voice queries in one of the largest search systems in Google. Finally, we see Mondegreen as complementing existing highly-optimized production ASR systems, which may not be frequently retrained and thus lag behind due to vocabulary drifts.

show abstract

“…Error correction has been applied in automatic speech recognition (ASR), which post-processes the outputs of the ASR system to achieve lower word error rate (WER) (Ringger and Allen, 1996;Cucu et al, 2013;D'Haro and Banchs, 2016;Tanaka et al, 2018). Taking the recognized sentence from the ASR system as source and the ground-truth sentence as target, ASR correction can be formulated as a sequence-to-sequence problem and modeled with autoregressive (Mani et al, 2020;Liao et al, 2020) or non-autoregressive (Leng et al, 2021) generation.…”

Section: Introductionmentioning

confidence: 99%

FastCorrect 2: Fast Error Correction on Multiple Candidates for Automatic Speech Recognition

Leng¹,

Tan²,

Wang³

et al. 2021

Findings of the Association for Computational Linguistics: EMNLP 2021

View full text Add to dashboard Cite

show abstract

Statistical Error Correction Methods for Domain-Specific ASR Systems

Cited by 18 publications

References 8 publications

BTS: Back TranScription for Speech-to-Text Post-Processor using Text-to-Speech-to-Text

BTS: Back TranScription for Speech-to-Text Post-Processor using Text-to-Speech-to-Text

Mondegreen: A Post-Processing Solution to Speech Recognition Error Correction for Voice Search Queries

FastCorrect 2: Fast Error Correction on Multiple Candidates for Automatic Speech Recognition

Contact Info

Product

Resources

About