Healthcare-SignNet: Advanced Medical Sign Language recognition using CNN + LSTM model

Sign Language Recognition (SLR) is essential for facilitating communication between the hearing- and speech-challenged community and those who do not understand sign language. This is especially essential within healthcare domain, where effective communication can impact diagnosis and patient care. This work introduces an innovative CNN+LSTM-based technique to SLR, designed specifically for healthcare applications. The Convolutional Neural Network (CNN) captures features of spatial nature from video frames, extracting critical elements such as shapes of hands, the movements, and facial expressions. These features are then processed by a Long Short-Term Memory (LSTM) network to extract and analyze the time-dependent variations in sign language sequences. The model is trained and tested on the curated dataset, which includes healthcare-specific sign language gestures. Achieving a recognition accuracy of 85.29%, our approach demonstrates state-of-the-art performance in healthcare focused SLR, offering a scalable and real-time solution to improve communication for patients with hearing impairments in medical settings.