Abstract: Computer Vision and deep learning techniques to recognize the hand gestures are among the trending domain of research now a days. The power of Artificial intelligence to improve the user interface and HCI is making human life much easier. Many researches are going on to develop systems that can understand hand gestures as input and perform several tasks. The communication through sign language is very ambiguous as it differs from person to person. This makes it very specific. Therefor, this project aims to build a system that can effectively determine a set of gestures, convert it to text and audio then perform certain task. At the same time it allows user to teach the system, their own gestures and associated messages to recognize.To accomplish this a CNN model is built to classify the gestures and Open CV is used for image capture and processing. After the model identifies the gesture it is converted to text/audio and associated task is performed.
Keywords: Computer Vision,Convolution Neural Network, Deep Learning,TensorFlow,Keras,Tkinter
| DOI: 10.17148/IJARCCE.2022.11118