✨👗 Aurelian’s Closet — Precision in Styling 👗✨
🌟 Inspiration
The inspiration for Aurelian’s Closet emerged from two seemingly unrelated observations.
🧵 First, fashion choices are one of the most personal and continuous forms of self-expression — evolving quietly with mood, routine, and life experience.
🧠 Second, through our background in computational neuroscience, we recognized that repeated everyday behaviors often contain measurable patterns that can reflect underlying cognitive states.
💡 We asked ourselves:
What if the most passive, universal form of self-expression—what we wear—could quietly reflect wellbeing?
What if choosing an outfit each morning told a deeper story than style alone?
This question led to Aurelian’s Closet — a platform that appears as an elegant luxury fashion experience on the surface, while discreetly embedding a powerful approach to passive, non-invasive health insight beneath.
👗✨ What It Does
🎭 On the Surface: A Precision Styling Experience
Aurelian’s Closet is a luxury-grade styling platform built on Bria FIBO’s structured JSON generation, offering:
- 🧥 Hyper-personalized outfit generation
- 🎥 Exact control over camera angles & FOV
- 💡 Professional lighting simulation
- 🧶 Detailed fabric texture & material control
- 🎨 Curated color palette engineering
- 🪞 AR-based virtual try-ons
- ☀️ Weather-aware, context-adaptive recommendations
All powered by FIBO’s professional controllability and reproducibility.
🔍 Beneath the Surface: Pattern Discovery
Quietly and ethically, the system:
- 📊 Tracks long-term style patterns
- 🎨 Monitors shifts in color and material preferences
- 🧵 Detects changes in outfit consistency and structure
- 📈 Correlates trends with established behavioral and neurological research
The result is the world’s first passive cognitive wellbeing insight system, seamlessly integrated into everyday dressing — no forms, no tests, no disruption.
🛠️ How We Built It
🧩 Technical Architecture
- 🖥️ Backend: Flask REST API (Python)
- 🎛️ AI Generation: Bria FIBO (JSON-native structured control)
- 🌐 Frontend: Responsive HTML / CSS / JavaScript
- 🗄️ Database: PostgreSQL for style history & pattern storage
- 🧠 Analysis Engine: PyTorch-based pattern recognition models
🎯 FIBO Integration Highlights
We built a full abstraction layer around FIBO to demonstrate its power:
- 📦 Structured JSON Control for every visual parameter
- 🎥 Professional Photography Controls (lighting, angles, depth)
- 🔁 Consistent & Reproducible Generation
- ⚙️ Scalable, Production-Ready Architecture
📊 Pattern Analysis System
- ⏳ Time-series analysis of style behavior
- 🔗 Correlation mapping without medical data
- 🕶️ Privacy-preserving, anonymized processing
- 📉 Visual insights without clinical terminology
🚀 Deployment
- 🐳 Dockerized infrastructure
- ✅ 95%+ test coverage
- 📚 REST API documentation
- 📓 Jupyter notebooks & sample datasets
⚠️ Challenges We Ran Into
🧑💻 Technical Challenges
- 🧠 Mastering FIBO’s full parameter space
- 🔬 Meaningful analysis without protected health data
- ⏱️ Optimizing real-time generation latency
- 🔐 Designing privacy-first analytics
🎨 Conceptual Challenges
- 🪄 Designing a dual-layer experience (fashion + insight)
- 🔄 Translating neuroscience research into fashion patterns
- ⚖️ Maintaining ethical, non-diagnostic outputs
🏆 Accomplishments We’re Proud Of
- 🥇 Complete FIBO Mastery in one cohesive system
- 🔀 Dual Innovation: Fashion generation + passive wellbeing insight
- 🏗️ Production-Ready Codebase in a hackathon timeframe
- 📈 Novel Pattern Discovery without sensitive data
- 👗 Elegant UX that hides deep complexity
- ✅ Professional Engineering Standards beyond typical hackathons
📚 What We Learned
🔧 Technical Learnings
- 📐 Structured generation beats prompt-only systems
- 📊 Pattern recognition thrives on longitudinal data
- 🔐 Privacy-first design is achievable and powerful
💡 Conceptual Learnings
- 👕 Daily behavior can act as a meaningful signal
- 🎭 Hidden complexity improves user trust
- 🔗 Breakthroughs happen at the intersection of disciplines
🏁 Hackathon Learnings
- 🎯 Focus beats feature overload
- 📖 Storytelling matters as much as tech
- 🎬 Build everything for the demo
🚀✨ What’s Next for Aurelian’s Closet
⏳ Short-Term (0–3 Months)
- 🏥 Research collaborations (with ethics approval)
- 🛍️ Fashion brand integrations
- 📱 iOS & Android apps with advanced AR
📆 Mid-Term (6–12 Months)
- 🧠 Expanded pattern recognition (stress, sleep, seasonality)
- 🧩 Enterprise-grade API
- 👥 Privacy-controlled community insights
🌍 Long-Term Vision
- 🌐 Global anonymized style-wellbeing dataset
- 🔮 Predictive, preventative health insights
- 👗 A fashion-health ecosystem built on trust
🧪 Technical Roadmap
- ⚙️ Deeper FIBO parameter control
- 🤖 Advanced ML pipelines
- ⌚ Wearable & smart-fabric integration
- 🔗 Blockchain-based data sovereignty
🎯 Impact Goals
- 🌈 Democratize passive wellbeing insight
- ♻️ Encourage sustainable, intentional fashion
- 🤝 Bridge fashion, technology, and healthcare
🧵✨ Closing
Aurelian’s Closet is more than a hackathon project.
It’s a new way of understanding how the smallest daily choices — like getting dressed — can quietly reflect who we are and how we’re doing.
One outfit at a time.
One pattern at a time.
A future where style and wellbeing move together. 💫
Built With
- alerting
- briafibo
- briafiboapi
- cloudrun
- cloudsql
- cloudstorage
- confluentcloud
- confluentkafkaapi
- css
- custommetricspipeline
- datadog
- datadogapi
- datadogapm
- docker
- dockercompose
- elevenlabs
- elevenlabssdk
- flask
- git
- googlecloudapi
- googlecloudplatform
- html
- insomnia
- javascript
- jupyternotebooks
- openweatherapi
- postgresql
- postman
- python
- pytorch
- redis
- sqlalchemy
- sqlite
- tensorflow
- vertexai
- visualstudiocode


Log in or sign up for Devpost to join the conversation.