DNN BASED GESTURE LANGUAGE TRANSLATOR

Gesture recognition is a powerful tool for creating interfaces that are adaptable to the needs of users. However, recognizing gestures can be challenging due to the inherent variability of human gestures. One application of gesture recognition is in sign language, which is used for communication and interaction between people. There are many different systems and methods for recognizing sign language, and our approach focuses on recognizing 26 symbols of hand motions used in the American Sign Language (ASL) for gesture recognition. The system is both sturdy and effective, allowing for real-time identification of static images classification by MK-ROD algorithm is proposed. The core idea behind a DNN-based gesture language translator is to use a sequence of input frames of a person making hand gestures and translate them into spoken language or written text. The procedure entails the utilization of a vast collection of annotated pairs of gestures and language to train a DNN model, which allows the model to acquire the ability to associate the two modalities. A success of DNN-based gesture language translators heavily relies on the quality, diversity of the training data and providing accuracy.