Luménoir

Luménoir transforms virtual try-on into cinematic self-discovery using AI-powered beauty personalization, aesthetic storytelling, and immersive visual identity experiences.

Comment

Inspiration

Modern beauty platforms are highly optimized for product discovery, but most virtual try-on experiences still feel transactional and repetitive. Users are often presented with generic filters or trend-driven recommendations that rarely feel personal or emotionally meaningful. I wanted to explore a different direction: what if AI-powered beauty technology could feel cinematic instead of commercial? Luménoir was inspired by the visual language of cinema, fashion editorials, and digital identity culture. Instead of simply helping users test products, the project focuses on helping users discover aesthetic moods and cinematic visual identities that align with their own facial structure, undertones, and style presence. The project was also inspired by the rise of creator culture and self-expression across digital platforms, where people increasingly use visual aesthetics as part of personal branding and identity.

What it does

Luménoir is an AI-powered cinematic identity experience built using Perfect Corp APIs and Gemini AI. Users upload a portrait image and select a cinematic aesthetic or visual mood. The platform then: Analyzes facial structure, undertones, and visual attributes using Perfect Corp APIs Applies AI-powered virtual try-on experiences Generates personalized cinematic styling interpretations Curates aesthetic-inspired beauty recommendations and visual themes Instead of generic beauty filters, Luménoir transforms virtual try-on into an immersive cinematic exploration experience. The goal is to make beauty technology feel: more artistic, more emotionally engaging, and more identity-driven.

How I built it

I developed Luménoir as a lightweight web application using a backend-focused architecture optimized for fast AI orchestration and smooth user interaction.

Frontend

The frontend was designed with Tailwind CSS using a cinematic luxury-inspired visual style featuring:

  • deep charcoal tones,
  • monochrome contrast,
  • soft lighting gradients,
  • and editorial-inspired layouts.

The goal was to make the experience feel closer to a cinematic identity studio than a traditional beauty application.

Backend

I built the backend using FastAPI to handle:

  • secure API communication,
  • asynchronous image processing,
  • file uploads,
  • and synchronization between multiple AI services.

This architecture allowed the application to remain lightweight, responsive, and modular during rapid iteration.

AI Integration

Luménoir integrates Perfect Corp’s AI-powered facial analysis and virtual try-on technologies to extract:

  • facial attributes,
  • skin undertones,
  • and visual styling compatibility.

The processed data is then passed into Gemini AI, which acts as the creative interpretation layer. Gemini transforms technical facial analysis into personalized cinematic aesthetic narratives and curated beauty styling recommendations.

This combination allows Luménoir to blend:

  • AI-driven beauty technology,
  • cinematic storytelling,
  • and immersive identity-focused experiences.

Challenges we ran into

As a solo developer working under tight hackathon time constraints, one of the biggest challenges was coordinating multiple AI-driven workflows while keeping the experience smooth and responsive.

API & Workflow Orchestration

The application required careful synchronization between:

  • image uploads,
  • external API processing,
  • asynchronous backend requests,
  • and dynamic frontend rendering.

Managing communication between Perfect Corp APIs, Gemini AI, and the frontend while avoiding slowdowns or request failures required constant testing and refinement.

Balancing Creativity with Usability

Another major challenge was balancing cinematic storytelling with practical usability. I wanted the AI-generated aesthetic narratives to feel artistic and immersive without overwhelming the user or distracting from the core virtual try-on experience.

Finding the right balance between:

  • visual atmosphere,
  • personalization,
  • and interface simplicity

became an important part of the design process.

Frontend Polish & Presentation

A significant amount of time also went into refining the visual presentation of the application. I focused heavily on:

  • monochrome contrast,
  • cinematic lighting aesthetics,
  • responsive layouts,
  • and maintaining a clean premium feel across different browsers and screen sizes.

Under hackathon time pressure, ensuring the application felt polished, stable, and consumer-facing was one of the most demanding parts of the project.

Accomplishments that I am proud of

I’m especially proud of transforming virtual try-on technology into something more emotionally immersive and creatively expressive. Rather than building a traditional beauty recommendation app, Luménoir explores how AI can support aesthetic identity discovery through cinematic storytelling and personalized visual interpretation. I’m also proud of building the project as a solo participant while integrating multiple APIs, designing the frontend experience, handling backend orchestration, and refining the overall product direction under limited time.

What I learned

This project taught me a great deal about:

  • AI orchestration,
  • asynchronous API workflows,
  • frontend presentation design,
  • and building consumer-focused AI experiences under tight time constraints.

Through the development process, I gained more practical experience working with:

  • image-based AI pipelines,
  • FastAPI backend architecture,
  • and coordinating communication between multiple external AI services.

One of the most valuable lessons was understanding how important clarity, presentation, and usability are when building AI-powered products. Even when the underlying technology is complex, the overall experience still needs to feel intuitive, visually polished, and easy for users to connect with.

As a solo developer, the project also helped me better understand the balance between technical implementation, design decisions, and creating an experience that feels engaging rather than overly complicated.

What's next for Luménoir

Looking beyond the hackathon, the next step for Luménoir is expanding the experience into a more interactive and personalized cinematic beauty platform.

Some future directions include:

  • Live AR Integration: Expanding beyond static image uploads into real-time AR-powered cinematic styling experiences using Perfect Corp’s live virtual try-on technologies.

  • Dynamic Aesthetic Personalization: Introducing more adaptive recommendations based on factors like lighting conditions, environment, visual mood, and user styling preferences.

  • Fashion & Accessory Pairing: Extending the platform beyond beauty into coordinated fashion, jewelry, and accessory recommendations that align with each user’s cinematic aesthetic profile.

  • Social Sharing & Creative Profiles: Allowing users to generate beautifully designed cinematic identity cards and share curated aesthetic looks across social platforms.

  • Retail & Product Discovery Integration: Exploring ways to connect curated beauty recommendations with real product discovery experiences and e-commerce integrations.

Longer term, I would like Luménoir to evolve into a broader AI-driven aesthetic discovery platform where virtual try-on, storytelling, personalization, and digital identity come together in a more immersive and expressive way.

Built With

Share this project:

Updates