Related Experiment Video
Updated: Jul 14, 2026
09:47
Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
Published on: December 15, 2023
922
Continuous Arabic Sign Language Recognition Models
Nahlah Algethami1, Raghad Farhud1, Manal Alghamdi1
1Computer Science Department, College of Computing and Informatics, Saudi Electronic University, Riyadh 11673, Saudi Arabia.
Sensors (Basel, Switzerland)
|May 14, 2025
Summary
This study introduces Temporal Convolutional Networks (TCN) for Arabic Sign Language (ArSL) recognition, achieving 99.5% accuracy. TCNs offer computational efficiency, bridging communication gaps for the deaf community.
Area of Science:
- Artificial Intelligence
- Computer Vision
- Natural Language Processing
Background:
- A significant communication gap exists between deaf and hearing communities, particularly for Arabic speakers due to limited Arabic Sign Language (ArSL) resources.
- Existing sign language recognition systems often lack the specialized datasets and models needed for ArSL.
Purpose of the Study:
- To develop and evaluate advanced deep learning models for accurate Arabic Sign Language recognition.
- To address the communication barriers faced by the Arabic-speaking deaf community by creating robust recognition systems.
Main Methods:
- Development of a custom dataset comprising the 30 most common ArSL sentences.
- Implementation and comparative analysis of Temporal Convolutional Network (TCN) and an enhanced Recurrent Neural Network (RNN) with Bidirectional Long Short-Term Memory (BiLSTM).
- Evaluation based on recognition accuracy, processing speed, and robustness to signing variations.
Main Results:
- The TCN model achieved a high accuracy of 99.5%.
- The enhanced RNN-BiLSTM model showed improvement from 96% to 99% accuracy.
- TCN demonstrated superior computational efficiency and faster inference times compared to the RNN-BiLSTM.
Conclusions:
- Temporal Convolutional Networks (TCN) are highly effective and computationally efficient for Arabic Sign Language recognition.
- The developed models show promise in bridging communication barriers for the hearing-impaired Arabic-speaking community.
- Further research into TCNs and enhanced RNNs can significantly advance sign language recognition technology.
