Indian Sign Language Recognition Using Deep Learning
Utkarsh Jagtap1, Vinayak Nangnurkar2, Suhas Chalwadi3, Neelam Jadhav4
Department of Computer Engineering
Genba Sopanrao Moze College of Engineering, Balewadi, Pune 45
------------------------------------------------------------***-----------------------------------------------------------
Abstract - This paper discusses the Implementation of sign language recognition systems, which is an important research area in the field of computer vision and machine learning. Indian Sign Language (ISL) serves as a crucial mode of communication for the hearing-impaired community in India. Recognizing and interpreting ISL gestures automatically is a challenging task due to the complex nature of the language. In this paper, we propose a novel approach for Indian Sign Language recognition using deep learning techniques.
The proposed system utilizes Mediapipe, a popular open-source framework, for extracting relevant hand and body features from video input. By leveraging the power of deep learning, we employ a Long Short-Term Memory (LSTM) network as the core model for prediction. LSTM's ability to capture temporal dependencies makes it well-suited for sign language recognition tasks.
The implementation process consists of several stages. Initially, a comprehensive dataset of ISL gestures is collected, annotated, and preprocessed. Mediapipe is then employed to extract key landmarks and features from the video sequences. The extracted features are fed into the LSTM network, which is trained on the dataset to learn the intricate patterns and dynamics of different sign gestures.
To evaluate the performance of our approach, we conducted extensive experiments using a standard evaluation protocol. The results demonstrate the effectiveness of the proposed system in recognizing ISL gestures accurately. Furthermore, we compare our approach with existing methods, showcasing its superiority in terms of recognition accuracy and robustness.The proposed Indian Sign Language recognition system holds significant potential for real-world applications, such as facilitating communication between hearing-impaired individuals and the general population. It has the ability to bridge the communication gap, promote inclusivity, and enhance the quality of life for the hearing-impaired community in India. Moreover, the methodology presented in this paper can serve as a foundation for future research in the field of sign language recognition, paving the way for advancements in other sign languages as well.
Keywords: Indian Sign Language, ISL, sign language recognition, deep learning, Mediapipe, Long Short-Term Memory, LSTM, feature extraction, gesture recognition