Despite the recent growth of online apparel shopping, customers are unable to physically try out a product which is probably one of the biggest cons of online shopping..
Moreover, regarding the present pandemic situation, where the spread of COVID-19 continues to dominate the public domain, people are afraid to go to real-shops in person.
Also, one often gets confused about how they can enhance their appearance and fashionability by doing small changes to their clothes or posture
What it does
we have made a intelligent fashion system in which we have 3 major feature:
Fashion recommender: its a model which tells us how with minimal adjustments to our clothes or posture , our looks can be more appealing there bby enhance our physical appearance
Make-up tryon: its a model through which the user or customer can virtually try on makeup upon their face give a reference face with make up on.
Clothes tryon: Its a model through which the user can virtually try on clothes on their body given a garment which thry want to try.
How we built it
we used GAN models for all the stated features using pytorch as our main tech stack. We referred to the SOTA research papers for developing our models.
Challenges we ran into
since the all these are still research topics, we underwent a hardtime to develop the models and make it work according to our needs. also since the datasets were huge, we faced problem while training our data as it required a lot of computational resources.
Accomplishments that we're proud of
we were able to make models according to our needs successfully and were satisfied with our results.
What's next for Intelligent Fashion System
we would like to add more features to it also make the web platform more interactive. since fashion recommender system is a completely new idea, we would like to develop more on that. also we will try to make a big platform for our idea.