PRupsa/Image-to-Caption-genrator

This code uses a pre-trained InceptionV3 CNN to extract image features, then feeds them through a Transformer encoder-decoder with attention and positional encoding for generating captions. It includes custom training and testing steps.

This repository is not active