American Sign Language Recognition Using Convolutional Neural Networks
Download datasets.rar and unzip them to datasets folder.
A lot of recent improvement has been made in the direction of developing computer vision systems that interpret sign language to spoken language. This technology often depends on on complex neural network architectures that can detect subtle patterns in streaming video. However, as a first step, towards understanding how to build a translation system, I have lessened the size of the problem by translating individual letters, instead of sentences. In this project, I have trained a convolutional neural network to classify images of American Sign Language (ASL) letters. After loading, examining, and preprocessing the data, I have trained the network and tested its performance.