Outfit Recommendation System Based on Deep Learning

Abstract

Abstract. In this paper, we propose an outfit recommendation system based on deep learning. Our goal is to use the system not only to judge an outfit if it is good or not but also to recommend good outfit to users when it is given a pool of cloth items. Our proposed model includes two parts: one is feature extractor based on ResNet-50, and the other is a binary classifier which is to classify the outfits into good ones and bad ones. Since our model is based on deep learning, it is necessary to use huge data to train the model. We collected a dataset which consists of 409,776 outfits with 644,192 items from the famous fashion website called Polyvore.com. With this dataset, we trained our model and the performance of it is over 84%. And our model can also recommend daily outfit to users.

Big data dressing recommendation based on ant colony algorithm to optimize BP network

“…To find the optimal fitness function is to find the corresponding weights and thresholds that can minimize the overall mean square error of the training set and the test set. The fitness function is formula (8):…”

An Efficient Style Virtual Try on Network for Clothing Business Industry

“…In GAN, the confrontation process of GAN training can detect unrealistic samples to ensure that they come from high-quality samples, which will also strongly confuse discriminators. It is widely used for visualizing what a network has learned [22], [23], [24], [25], [26], and recently to synthesize images [27], [28].In particular, [27] also generates clothing images, but they generate single-garment products rather than full body out-fits.…”