Inspiration
We both had a really strong interest in Ai, so we decided that neural networks would be in both of our interests. In our time working on the project we faced challenges in accuracy testing where the OCR wouldn't accurately identify the image file based on our parameters.
What it does
It takes a noisy image with text that and then runs an OCR model to detect text then outputs it with 96.69% accuracy. It then goes to stage 2 which compresses the output text using adaptive Huffman encoding which can be later decompressed.
Log in or sign up for Devpost to join the conversation.