Year: 2021
Type: conference-paper
Abstract: Although automatic speech recognition (ASR) systems achieved significantly improvements in recent years, spoken language recognition error occurs which can be easily spotted by human beings. Various language modeling techniques have been developed on post recognition tasks like semantic correction. In this paper, we propose a Transformer based semantic correction method with pretrained BART initialization, Experiments on 10000 hours Mandarin speech dataset show that character error rate (CER) can be effectively reduced by 21.7% relatively compared to our baseline ASR system. Expert evaluation demonstrates that actual improvement of our model surpasses what CER indicates.
Source: Interspeech (USB)
Authors: Yun Zhao, Xuerui Yang, Jinchao Wang, Yongyu Gao, Chao Yan, Yuanfu Zhou
Institution: Arizona State University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W3152575505