Real Time Hand Gesture Recognition Using LSTM Based Deep Learning
Vijaya Kumar Gurrala, Srinivas Talasila, Vaishnavi R C, J. Shruthi, Jessu Supreeth · 2025
This research introduces a cutting-edge system for real-time hand gesture recognition using deep learning techniques. Gesture recognition is a vital technology that can enhance numerous applications, including sign language translation, secure communications, and various healthcare services. In this methodology, implementation of Long Short-Term Memory (LSTM) networks, a specialized type of Recurrent Neural Network (RNN) is observed, which is well suited for sequential data analysis. Initially, the captured dataset undergoes data augmentation. The proposed system operates by capturing live video feeds to extract essential key points from hand movements, which are then analyzed to recognize specific gestures through a sophisticated deep-learning algorithm. By concentrating on the spatial and temporal dynamics of hand movements, this approach significantly improves both accuracy and responsiveness in real-time interactions. The overarching aim of this project is to facilitate effective communication between children who are deaf and mute and their parents or caregivers. By encouraging the learning and use of sign language, this gesture recognition system aims to enhance social interaction for these children. Furthermore, the system functions as an intermediary, seamlessly converting sign language gestures into spoken or written text in real-time. The proposed method demonstrates exceptional performance, achieving an accuracy of 99.6%. The results validate the robustness of the implemented techniques and suggest that the methodology could be highly beneficial for real-world applications.