Developed a mobile AR application using Unity that place 3D assets in real-world environments and capture structured scene data for downstream AI pipelines.
Project Year
2025
Softwares/Frameworks Used
Unity (AR Foundation) , ARKit (iOS) , C#, Backend APIs (Cloud , Signed URL workflows)
AR Interaction System
Built a mobile AR application using Unity (AR Foundation with ARKit support)
Enabled users to:
Upload and place 3D models (.glb) in real-world environments
Interact with assets through placement, repositioning, and scene control
Implemented plane detection and surface tracking for stable AR placement
Structured Data Capture
Designed capture pipeline to extract structured outputs from AR sessions:
Background image (camera feed)
Foreground mask (binary alpha)
HDR environment data (CubeMap / lighting information)
Ensured captured data is suitable for downstream processing and compositing workflows
Backend Integration
Integrated with backend services using signed URL uploads
Designed payload system for transmitting AR session data to processing pipelines
Enabled asynchronous processing workflows with real-time feedback in the app
Multi-Modal Processing Workflows
Supported both:
3D workflows → environment-aware processing using captured spatial data
2D workflows → image-based processing using captured frames
Enabled flexible pipeline integration depending on downstream requirements
Performance & Optimization
Implemented image compression for faster data transfer
Optimized model handling for mobile performance constraints
Ensured stable AR tracking and responsive interaction on supported iOS devices
Use Case & Impact
Bridges real-world capture with digital content creation pipelines
Enables scalable data acquisition for AI-driven and rendering workflows
Supports rapid iteration and content generation from user-driven inputs