A Web-Based Audio-to-Sign Language Converter for Chinese National Sign Language

Document Type

Conference Proceeding

Publication Date

1-1-2022

Abstract

About 20.54 million people in China suffer from hearing impairment, accounting for a quarter of the disabled population. However, only hundreds of thousands of people in China have revived formal sign language training. Therefore, for hearing-impaired people, how to communicate with ordinary people becomes a problem. Some articles have looked at converting sign language into text or audio. However, very few articles have dealt with how to help hearing-impaired people understand what ordinary people say. Therefore, this paper proposes a new method to convert audio into sign language video and play it on the web page. The main idea is to build a database based on Chinese common sign language and collect audio through computer microphones. After that, this method uses the Chinese word segmentation method of HanLP to segment words, and the obtained phrases are processed and matched with the database. After obtaining the database output, the videos are sorted and spliced by matching the original phrase list. Eventually, the stitched video will be played on the web. Experimental results show that the accuracy of this method is 90%. The relative code can be found at https://github.com/MelanthaWang246/A2SL-CN.

Publication Title

2022 4th International Academic Exchange Conference on Science and Technology Innovation, IAECST 2022

First Page Number

1045

Last Page Number

1048

DOI

10.1109/IAECST57965.2022.10061950

This document is currently not available here.

Share

COinS