Input Length Matters: Improving RNN-T and MWER Training for Long-form Telephony Speech Recognition

no code implementations • 8 Oct 2021 • Zhiyun Lu, Yanwei Pan, Thibault Doutre, Parisa Haghani, Liangliang Cao, Rohit Prabhavalkar, Chao Zhang, Trevor Strohman

Our experiments show that for both losses, the WER on long-form speech reduces substantially as the training utterance length increases.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +1

Bridging the gap between streaming and non-streaming ASR systems bydistilling ensembles of CTC and RNN-T models

no code implementations • 25 Apr 2021 • Thibault Doutre, Wei Han, Chung-Cheng Chiu, Ruoming Pang, Olivier Siohan, Liangliang Cao

To improve streaming models, a recent study [1] proposed to distill a non-streaming teacher model on unsupervised utterances, and then train a streaming student using the teachers' predictions.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +1

Improving Streaming Automatic Speech Recognition With Non-Streaming Model Distillation On Unsupervised Data

no code implementations • 22 Oct 2020 • Thibault Doutre, Wei Han, Min Ma, Zhiyun Lu, Chung-Cheng Chiu, Ruoming Pang, Arun Narayanan, Ananya Misra, Yu Zhang, Liangliang Cao

We propose a novel and effective learning method by leveraging a non-streaming ASR model as a teacher to generate transcripts on an arbitrarily large data set, which is then used to distill knowledge into streaming ASR models.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +1