A Digital Sign Language Recognition based on a 3D-CNN System with an Attention Mechanism

Sign language recognition makes communication between hearing impaired people and people more convenient. The rapid development of deep learning has brought new opportunities for sign language recognition. This paper proposes a 3D-CNN with the attention mechanism for real-time digital sign language recognition. The background noise in the gesture data extraction process is removed, and the introduction of the attention mechanism enables the model to enhance the feature expression from the angle information. The experimental results show that our system performs real-time recognition of digital sign language well, and the accuracy rate reaches 98.49%.