Summary:
Our main idea revolves around helping deaf people to understand sign language without learning it. Simply by performing such signs and being interpreted and translated to letters and words can help people understand it with high ease. The main population this technology can help are the people suffering difficulty in learning sign language. This solution can be applied almost everywhere including educational institutions, government offices to commercial places. This can help in easy learning and working activities. Also, including future possible developments during seminars and other public speaking events, real-time prediction systems can provide a very efficient method of delivery. Hand signs in general, follow the following structure based on which, our model will be trained, and outputs will be predicted.
I have tested with the follwing models
- CNN
- LeNet-5