Inspiration

Buying clothes online is frustrating when generic size charts fail. We wanted to eliminate fit guesswork and bring a personalized physical fitting room experience straight to your smartphone screen.

What it does

Our AI platform transforms two photos into a virtual fitting room in seconds. It extracts precise body measurements, maps complementary color palettes, and renders dresses tailored onto your exact body shape. To ensure accuracy and transparency, the system provides a confidence percentage for every measurement and color assessment before generating the final render.

How we built it

  • Core AI Reasoning Engine: Powered by Gemini, which drives our multimodal visual analysis to extract body proportions, evaluate skin undertones with confidence metrics, and coordinate color matching.
  • Image and Video Orchestration: Integrated Vonage APIs to handle high-reliability user image capture, video ingestion, and media delivery across channels seamlessly.
  • Agent Infrastructure and White-Labeling: Leveraged Google Antigravity as our underlying agent framework to execute complex computer vision workflows, manage multi-step processing pipelines, and white-label the entire solution for seamless brand integration.

Challenges we ran into

Mapping complex garment fabric drapes accurately onto diverse body shapes while keeping rendering times low presented major computational and spatial alignment hurdles, particularly when balancing automated measurement scoring with rendering fidelity.

Accomplishments that we're proud of

We successfully automated the end-to-end pipeline from initial photo upload to color-matched, photorealistic image generation, complete with real-time confidence scores for fit accuracy.

What we learned

Integrating computer vision with generative AI requires precise image preprocessing to maintain garment fidelity, color accuracy, and anatomical proportions across varying lighting conditions and body types.

What's next for Dress Me

We plan to expand our intelligence capabilities by introducing automated fabric texture recommendations based on climate and style preference, introducing real-time shipment and order tracking APIs, and scaling beyond single dresses into full multi-garment layering.

Built With

Share this project:

Updates

Submission history