IMAGE-BASED RECOMMENDATION USING RESNET

Abstract— The goal of this project is to provide product recommendations using image search. The recommendations will be provided on the basis of image similarity using the ResNet50 model. The ResNet model is a technique that uses deep learning to recommend items based on their visual similarity. The system allows users to upload an image of a product and then retrieves similar products from a database based on the visual features extracted from the image using the ResNet50 model. The nearest neighbor algorithm is used to find the most similar products to the uploaded image based on the Euclidean distance between the feature vectors. The top five most similar products are returned and displayed to the user.