Hybrid model for fashion recommendation system using image by CNN and transformer techniques

Fashion is viewed as a desirable and lucrative industry since fashionable products are in high demand. One of the features that will ensure customer comfort is the ability to recommend similar fashion items based on the image of fashion item. This feature fetches similar fashion items for the customer based on the fashion image given as an input by the customer instead of searching with qualitative data which will consume more time compared to searching with images. To classify and recommend fashion item, this research involves multi label classification of fashion items by applying convolution-based algorithms like CNN, RESNET 50 and vision transformer-based algorithm such as Swin transformer and recommendation algorithm to recommend fashion items and also showcases the integration of trained model.