Recognizing American Sign Language gestures efficiently and accurately using a hybrid transformer model

Abstract Gesture recognition plays a vital role in computer vision, especially for interpreting sign language and enabling human–computer interaction. Many existing methods struggle with challenges like heavy computational demands, difficulty in understanding long-range relationships, sensitivity to...

Full description

Saved in:
Bibliographic Details
Main Authors: Mohammed Aly, Islam S. Fathi
Format: Article
Language:English
Published: Nature Portfolio 2025-06-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-06344-8
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849433113838485504
author Mohammed Aly
Islam S. Fathi
author_facet Mohammed Aly
Islam S. Fathi
author_sort Mohammed Aly
collection DOAJ
description Abstract Gesture recognition plays a vital role in computer vision, especially for interpreting sign language and enabling human–computer interaction. Many existing methods struggle with challenges like heavy computational demands, difficulty in understanding long-range relationships, sensitivity to background noise, and poor performance in varied environments. While CNNs excel at capturing local details, they often miss the bigger picture. Vision Transformers, on the other hand, are better at modeling global context but usually require significantly more computational resources, limiting their use in real-time systems. To tackle these issues, we propose a Hybrid Transformer-CNN model that combines the strengths of both architectures. Our approach begins with CNN layers that extract detailed local features from both the overall hand and specific hand regions. These CNN features are then refined by a Vision Transformer module, which captures long-range dependencies and global contextual information within the gesture. This integration allows the model to effectively recognize subtle hand movements while maintaining computational efficiency. Tested on the ASL Alphabet dataset, our model achieves a high accuracy of 99.97%, runs at 110 frames per second, and requires only 5.0 GFLOPs—much less than traditional Vision Transformer models, which need over twice the computational power. Central to this success is our feature fusion strategy using element-wise multiplication, which helps the model focus on important gesture details while suppressing background noise. Additionally, we employ advanced data augmentation techniques and a training approach incorporating contrastive learning and domain adaptation to boost robustness. Overall, this work offers a practical and powerful solution for gesture recognition, striking an optimal balance between accuracy, speed, and efficiency—an important step toward real-world applications.
format Article
id doaj-art-caccf491ef1b4e53a3bb2789e2f0887a
institution Kabale University
issn 2045-2322
language English
publishDate 2025-06-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-caccf491ef1b4e53a3bb2789e2f0887a2025-08-20T03:27:10ZengNature PortfolioScientific Reports2045-23222025-06-0115112710.1038/s41598-025-06344-8Recognizing American Sign Language gestures efficiently and accurately using a hybrid transformer modelMohammed Aly0Islam S. Fathi1Department of Artificial Intelligence, Faculty of Artificial Intelligence, Egyptian Russian UniversityDepartment of Computer Science, Faculty of Information Technology, Ajloun National UniversityAbstract Gesture recognition plays a vital role in computer vision, especially for interpreting sign language and enabling human–computer interaction. Many existing methods struggle with challenges like heavy computational demands, difficulty in understanding long-range relationships, sensitivity to background noise, and poor performance in varied environments. While CNNs excel at capturing local details, they often miss the bigger picture. Vision Transformers, on the other hand, are better at modeling global context but usually require significantly more computational resources, limiting their use in real-time systems. To tackle these issues, we propose a Hybrid Transformer-CNN model that combines the strengths of both architectures. Our approach begins with CNN layers that extract detailed local features from both the overall hand and specific hand regions. These CNN features are then refined by a Vision Transformer module, which captures long-range dependencies and global contextual information within the gesture. This integration allows the model to effectively recognize subtle hand movements while maintaining computational efficiency. Tested on the ASL Alphabet dataset, our model achieves a high accuracy of 99.97%, runs at 110 frames per second, and requires only 5.0 GFLOPs—much less than traditional Vision Transformer models, which need over twice the computational power. Central to this success is our feature fusion strategy using element-wise multiplication, which helps the model focus on important gesture details while suppressing background noise. Additionally, we employ advanced data augmentation techniques and a training approach incorporating contrastive learning and domain adaptation to boost robustness. Overall, this work offers a practical and powerful solution for gesture recognition, striking an optimal balance between accuracy, speed, and efficiency—an important step toward real-world applications.https://doi.org/10.1038/s41598-025-06344-8Gesture recognitionSign language recognitionHybrid transformer-CNNDeep learningReal-time inference
spellingShingle Mohammed Aly
Islam S. Fathi
Recognizing American Sign Language gestures efficiently and accurately using a hybrid transformer model
Scientific Reports
Gesture recognition
Sign language recognition
Hybrid transformer-CNN
Deep learning
Real-time inference
title Recognizing American Sign Language gestures efficiently and accurately using a hybrid transformer model
title_full Recognizing American Sign Language gestures efficiently and accurately using a hybrid transformer model
title_fullStr Recognizing American Sign Language gestures efficiently and accurately using a hybrid transformer model
title_full_unstemmed Recognizing American Sign Language gestures efficiently and accurately using a hybrid transformer model
title_short Recognizing American Sign Language gestures efficiently and accurately using a hybrid transformer model
title_sort recognizing american sign language gestures efficiently and accurately using a hybrid transformer model
topic Gesture recognition
Sign language recognition
Hybrid transformer-CNN
Deep learning
Real-time inference
url https://doi.org/10.1038/s41598-025-06344-8
work_keys_str_mv AT mohammedaly recognizingamericansignlanguagegesturesefficientlyandaccuratelyusingahybridtransformermodel
AT islamsfathi recognizingamericansignlanguagegesturesefficientlyandaccuratelyusingahybridtransformermodel