FitMii

Online shopping has taken its boom in the past two decades, however, there is still no proper way to gauge the correct fitting for the buyer i.e. how the clothing would look on the wearer, and the proper size since everyone has different body dimensions. It can be really hard to estimate the proper

2025-06-28 16:27:16 - Adil Khan

Project Title

FitMii

Project Area of Specialization Augmented and Virtual RealityProject Summary

Online shopping has taken its boom in the past two decades, however, there is still no proper way to gauge the correct fitting for the buyer i.e. how the clothing would look on the wearer, and the proper size since everyone has different body dimensions. It can be really hard to estimate the proper fitting of clothing from the images featured on clothing websites, they usually have a model or plastic doll wearing the clothes, both of them are not the true representation of the general public.

There are virtual avatars available on the internet but they do not take into account the proper measurements of the human body, and are pretty inaccurate. We will be using various techniques and methods to accomplish our goal. Python will be the main language in focus, as it allows for easy web scrapping. FitMii will be integrated with on cloud GPU to render out a low-poly 3D model based on user measurements, this model will either be made using Blender/Unity/Unreal Engine. The same GPU would also be used for machine learning and detecting clothes from online articles. Our solution will take the clothing articles from the supposed website, and then process them into usable textures, that can be later mapped on a virtual avatar. This avatar will be based on the actual measurements of the user. Providing the user with an estimated look with clothes based on their body type. 

Project Objectives
  1. To create a 3D model of human body based on user's measurements which would be taken via AR methods or user defined. 
  2. To detect clothing apparels like shirts from the images.
  3. To carefully cut-out those clothing articles and convert them into textures.
  4. To map that texture onto the 3D model created earlier
Project Implementation Method

FDD (Feature Driven Development) will be used for this project. It is mainly because one feature will be linked to another one in the future, therefore this approach suits it best.  

Benefits of the Project

The beneficiaries of this project are the users, and clothing brands since they will have a more realistic way of representing clothing items to people. Moreover, the work for converting clothing items into textures rapidly can be used in the gaming/metaverse field as well.

Technical Details of Final Deliverable

The final deliverable would be a prototype for 'FitMii' that would make use of server-side GPU to extract clothing items from images and convert them into textures, and finally map them on 3D model.

Final Deliverable of the Project Software SystemCore Industry ManufacturingOther Industries Others Core Technology Augmented & Virtual RealityOther Technologies Artificial Intelligence(AI)Sustainable Development Goals Industry, Innovation and InfrastructureRequired Resources
Item Name Type No. of Units Per Unit Cost (in Rs) Total (in Rs)
Total in (Rs) 50000
GPU Equipment15000050000

More Posts