This project was done under Amazon ML challenge 2024. The task was to build and train a Deep learning model which can analyze an image of a product and extract the product's features such as weight , dimensions, volumne, volatge, watts, etc (if lablled on the inuput image).

After trying and experimenting out with several OCR libraries we decided to go with PeddleOCR for its better accuracy and perfromance on tilted texts in the image.