Published Paper Details:

Sign Language Recognition (SLR) targets interpreting the sign language into text or speech, to facilitate the communication between deaf-mute people and ordinary people. This task has a broad social impact but is still very challenging due to the complexity and large variations in hand actions. Existing methods for SLR use hand-crafted features to describe sign language motion and build classification models based on those features. However, it is difficult to design dependable features that can adapt to the wide range of hand gestures. To address this issue, we propose a novel convolutional neural network (CNN) that automatically extracts discriminative spatial-temporal features from raw video streams without any prior knowledge, thereby avoiding feature design. Multi-channel video streams with color information, depth clues, and body joint positions are used as input to the CNN to integrate color, depth, and trajectory information to improve performance. On a real dataset gathered with Microsoft Kinect, we approve the proposed model and show how it beats conventional methodologies in light of hand-created highlights. The created framework will be utilized as a learning device for gesture-based communication fledglings including hand recognition. The pictures were taken care of into a model called the Convolutional Neural Network for order (CNN). Keras was utilized to prepare the pictures. A uniform foundation and legitimate lighting are given. The's undertaking will likely foster an AI model that can order the different hand motions utilized in gesture-based communication fingerspelling.