ArtSpace

website ver available but no AR feature.

Inspiration

We were inspired by the idea of seeing classical artworks like the Mona Lisa in our room, knowing that most people can't own such masterpieces. But beyond famous art, we realized that 70% of online art buyers hesitate to purchase because they can't visualize how a piece will look in their actual space. We wanted to bridge this gap between digital browsing and physical reality, making art accessible and tangible for everyone.

What it does

ArtSpace is an AR-powered virtual art gallery that lets users visualize artworks in their real environment before purchasing. Users can browse 40+ curated artworks (both 2D paintings and 3D sculptures), view them at real-world scale using augmented reality, and pinch-to-scale to resize pieces to fit their space perfectly. Our AI Advisor recommends artworks based on user preferences, and buyers can instantly chat with sellers. Users can also upload their own 3D models or generate them from photos using AI.

How we built it

We built ArtSpace using pure web technologies for maximum accessibility, our stack includes:

HTML5, CSS3, JavaScript for the responsive web interface
Google Model-Viewer for 3D rendering and AR visualization
WebXR API for cross-platform AR support (iOS Safari & Android Chrome)
Meshy.ai API for AI-powered image-to-3D model generation
GLB/GLTF format for optimized 3D models
GitHub Pages for deployment ## Challenges we ran into
iOS Safari compatibility - File upload buttons weren't responding on iOS; we had to restructure our HTML using label elements and transparent input overlays
Storage limitations - Base64-encoded 3D models exceeded localStorage's 5MB limit, causing "storage full" errors. We pivoted to blob URLs with sessionStorage
AR scale persistence - Initially, resized models would snap back to original size. We changed from ar-scale="fixed" to ar-scale="auto" to allow pinch gestures to persist
3D model loading - Large GLB files caused timeout issues; we implemented loading indicators and error handling for better UX ## Accomplishments that we're proud of
Pinch-to-scale AR where resized models stay fixed - users can perfectly fit art to their walls
AI-powered 3D generation - users can photograph objects and convert them to AR-viewable 3D models
40+ curated artworks with real dimensions, pricing, and seller contact information
Seamless UX with shortlisting, comparison tools, and an AI advisor chatbot ## What we learned
Mobile browser quirks (especially iOS Safari) require extensive testing and workarounds
Storage management is critical for web apps handling large binary files
User experience in AR requires careful consideration of scale, placement, and interaction feedback ## What's next for ArtSpace
Gallery partnerships - Onboarding real galleries and artists to sell through our platform
Social features - Allow users to share AR screenshots of art in their space
Virtual exhibitions - Curated AR gallery walks users can experience in their homes