Skip to content

Shoes & Bags AR Try-On

Real-time 3D/AR virtual try-on solution for footwear, bags, and backpacks.

What It Does

3D product models are displayed live on the camera view (augmented reality) allowing shoppers to virtually try on shoes, bags, and backpacks before purchasing. The system uses machine learning for foot/body detection and optical flow for smooth, stable tracking.

Key Features

  • Mobile AR — point your phone camera at your feet to try on shoes in real-time
  • Desktop viewer — on-screen try-on experience for monitors
  • Mirror mode — designed for large digital mirrors in physical retail stores (mirror modes 1–5)
  • 3D product management — upload GLB models, edit textures and materials, adjust lighting, position products
  • Tron editor — fine-tune product positioning in the AR scene with ML-assisted auto-positioning
  • Auto-positioning — shoes are automatically positioned using ML; bags/backpacks require manual setup
  • Turbo mode — alternating ML/optical-flow pipeline for higher FPS on footwear
  • Pose detection — multiple quality levels (lite/full/heavy) for accessories on body (bags, backpacks)
  • Multi-language — UI supports English, Polish, German, Czech, Slovak, Lithuanian, Latvian, and more
  • Video recording — capture try-on sessions as MP4 (via settings panel)
  • Partner white-labeling — custom branding for partners (e.g., Zara, Inditex, Nanovo)

Integration Options

Method Description Best for
AR code / link Client gets a code or URL to put on their website Simplest setup
Iframe embed Embed the try-on viewer in an iframe on any page Seamless web integration
Digital mirror Full-screen AR on large display with camera Physical retail stores
Shopify plugin Native integration via the WEARFITS Shopify App Shopify stores
REST API Programmatic model upload and management Custom integrations

Domain

Adding Products

Products can be added via: 1. Web interface at dev.wearfits.com — upload GLB/GLTF/OBJ/STL files, manage models visually 2. REST API — programmatic product management (see API Reference) 3. 2D→3D Generator — if you don't have 3D models, use the Shoe 2D→3D converter to auto-generate them from photos

Supported File Formats

  • Upload: .glb, .gltf, .obj, .stl, .zip (containing OBJ + MTL + textures)
  • Export: GLTF, OBJ, DRC (Draco compressed), USDZ (for Apple AR)
  • Max file size: 30 MB per upload

Tech Stack

Component Technology
Runtime Node.js 24.x on GCP App Engine (Flexible)
Storage GCP Cloud Storage (3D objects, textures, configs)
ML TensorFlow.js, ONNX Runtime, native WASM models
3D Rendering Three.js
Real-time Socket.io

Further Reading