American Sign Language (ASL) letter classification with pytorch.
This is a Multiclass image classification with CNN on American Sign Language Dataset.
The dataset involves images from the American Sign Language which were seprated into 29 different folders. Each folder represented a different class.
The classes are made up of twenty-six letters (A-Z) and three classes for 'SPACE', 'DELETE', and 'NOTHING'. In all, the training set contains 87,000 images whuch are 200x200 pixesls in width and height.