Logo image
Back
Text Injection for Neural Contextual Biasing
Preprint   Open access

Text Injection for Neural Contextual Biasing

Zhong Meng, Zelin Wu, Rohit Prabhavalkar, Cal Peyser, Weiran Wang, Nanxin Chen, Tara N Sainath and Bhuvana Ramabhadran
arXiv (Cornell University), p.5 pages
06/11/2024
DOI: 10.48550/arxiv.2406.02921
url
https://doi.org/10.48550/arXiv.2406.02921View
Preprint (Author's original)This preprint has not been evaluated by subject experts through peer review. Preprints may undergo extensive changes and/or become peer-reviewed journal articles. Open Access

Abstract

Neural contextual biasing effectively improves automatic speech recognition (ASR) for crucial phrases within a speaker's context, particularly those that are infrequent in the training data. This work proposes contextual text injection (CTI) to enhance contextual ASR. CTI leverages not only the paired speech-text data, but also a much larger corpus of unpaired text to optimize the ASR model and its biasing component. Unpaired text is converted into speech-like representations and used to guide the model's attention towards relevant bias phrases. Moreover, we introduce a contextual text-injected (CTI) minimum word error rate (MWER) training, which minimizes the expected WER caused by contextual biasing when unpaired text is injected into the model. Experiments show that CTI with 100 billion text sentences can achieve up to 43.3% relative WER reduction from a strong neural biasing model. CTI-MWER provides a further relative improvement of 23.5%.
 Computer Science - Artificial Intelligence Computer Science - Computation and Language Computer Science - Learning Computer Science - Neural and Evolutionary Computing

Details

Metrics

9 Record Views
Logo image