Draculothes — Your Digital Wardrobe, Without a Mirror

Inspiration

Have you ever wondered how Dracula gets ready for a party when he can't see his reflection?

Inspired by the Halloween theme of helping fictional characters solve everyday problems, we chose Dracula as our character. As the host of Hotel Transylvania, he frequently attends and hosts parties, meaning he needs different outfits for different occasions. However, he faces two problems: he cannot see his reflection in mirrors, and he cannot easily visit clothing stores during the day.

This inspired us to create Draculothes, a virtual fitting room that lets Dracula see how different outfits look without needing a physical mirror.

What It Does

Draculothes is an AI-powered virtual wardrobe that allows users to create a 3D avatar of themselves and virtually try on clothes.

Users can:

Create a 3D avatar using a full-body photo.
Browse clothing from online stores.
Turn clothing images into 3D models.
Mix and match clothes on their avatar.
View their outfits from different angles.

Our goal is to make online shopping more interactive and convenient.

How We Built It

We designed Draculothes around three main steps:

Create a Profile: Users upload a full-body photo, which is converted into a 3D avatar using Tripo3D.
Find Clothes: Users browse online stores and select clothing items. Tripo3D is used to generate 3D clothing models from images.
Mix and Match: The 3D avatar and clothing models are brought together to create a virtual try-on experience.

For future development, we plan to use Blender and Mixamo to rig the avatar and fit the clothes, Three.js to display and animate the models, and MediaPipe to track users' body movements through their webcams.

Challenges We Faced

One of our biggest challenges was figuring out how to turn regular images into usable 3D models. Generating a 3D model is only the first step; making clothing fit naturally onto a human avatar requires additional modelling and rigging.

We also had to explore how different technologies could work together. Tools like Tripo3D, Blender, Mixamo, and Three.js each serve different purposes, so connecting the entire workflow was another challenge.

Since we were working with several unfamiliar technologies, understanding their capabilities and limitations was an important part of developing our idea.

What We Learned

Throughout this project, we learned more about AI-powered 3D generation, 3D modelling, and how interactive web applications work.

We also discovered that generating 3D objects and making them interact with each other are two very different challenges. Beyond coding, we learned how to break down a complex idea into smaller components and explore different tools to bring them together.

Most importantly, we learned that even fictional characters can inspire real-world solutions.