Human gestures are a non-verbal method of communication that is essential in interactions between humans and robots. In order to recognize hand movements and facilitate such interactions, visionbased gesture recognition techniques are crucial. A simple and useful interface between gadgets and people is made possible by hand gesture recognition. Hand gestures may be employed in many different contexts, making them useful for communication and other purposes. People with hearing loss or disabilities, as well as those who have had strokes, might benefit from hand gesture recognition because they need to be able to interact with others by employing gestures that are universally understood, such as the signs for food, drink, family, and more. This study suggests a method for identifying hand motions using convolutional neural networks (CNN). Based on a number of criteria, including execution time, accuracy, sensitivity, specificity, positive and negative predictive value, probability, and root mean square, the proposed approach is assessed and contrasted between training and testing modes. The results demonstrate that CNN is a successful method for identifying distinctive characteristics and categorizing data, with testing accuracy of 100%

Computer vision, Machine learning, Deep learning, Pattern recognition, Hand gesture, Convolutional neural network