Artificial intelligence based real time deciphering of British sign language

This paper presents an approach to detect sign language to establish means of communication with mute and deaf people. Sign language is a method which helps mute and blind people to overcome the communication barrier. The proposed system uses a Long Short-Term Memory (LSTM) model with Mediapipe holi...

Full description

Saved in:
Bibliographic Details
Main Authors: Shilaskar, Swati, Bhatlawande, Shripad, Singh, Aaditya, Andhare, Niharika, Singh, Aditya, Kumar, Aditya, Madake, Jyoti
Format: Conference Proceeding
Language:English
Subjects:
Artificial intelligence
Datasets
Real time
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper presents an approach to detect sign language to establish means of communication with mute and deaf people. Sign language is a method which helps mute and blind people to overcome the communication barrier. The proposed system uses a Long Short-Term Memory (LSTM) model with Mediapipe holistics which detects landmarks on hand for detection of hand gestures which helps in recognizing and deciphering British sign language (BSL) in Real Time. In BSL both hands are used for gesturing. The dataset is created manually by the authors. This resulted in a responsive and accurate real time detection of BSL. The proposed method is simple with limited calculations and works well with a smaller dataset. The real time detection of the proposed method has an accuracy of 89%.
ISSN:0094-243X
1551-7616
DOI:10.1063/5.0179287