Nadipally Srinibha Rao

Product Recommendation System

The rise in living standards has fueled a growing interest in fashion as a prominent aesthetic expression. The abundance of clothing options online has made it challenging for customers to find the right outfit. However, conventional systems relying on user history fall short, and our project seeks to innovate by utilizing user-input images to generate tailored recommendations. We have used a pre-trained model(ResNet50 trained on ImageNet) to extract features from a fashion dataset and then used a classifier K-Nearest Neighbour algorithm and cosine similarity to build our recommender system.