Computer Vision project for hand gesture recognition of the American Sign Language
Hand gesture recognition is the process of recognizing meaningful expres- sions of form and motion by a human involving only the hands. There are plenty of applications where hand gesture recognition can be applied for improving control, accessibility, communication and learning. In this work, we present our results in classifying twenty six hand signs for the letters of the alphabet. We conduct experiments using different convolutional neural networks architectures and report the performance of each of them.