A Real use case of practical AR-AI couple: Size my luggage
summary
A working Android prototype uses ARCore + Sceneform to place a transparent 3D suitcase model and Firebase ML Kit object detection to find a real bag’s bounding box; the app runs on-device, pulls renderables from assets or web, and lets users drag/rotate (no scale) to visually compare sizes. It requires good lighting and textured floors, ML Kit returns only coarse labels so false positives occur, and library conflicts needed dependency exclusions; source and demo code sit on GitHub.
AI-written summary · about
chapters (8)
- 0:00Introduction and background
- 4:13AR use cases in travel
- 6:40ARCore framework overview
- 10:14Sceneform for AR development
- 13:08Building the luggage AR app
- 23:42Enhancing with AI and MLKit
- 34:21Live demo and integration
- 38:17Q&A and conclusion
related talks
37mA Real use case of practical AR-AI couple - Size my luggage
41mFusing AI with AR: Custom Markers
39mNow smile (also in 3D)! Exploring AR, ML and Camera related APIs on Android
30mMachine Learning in Android Apps with ML Kit
18mDeep-dive on ARCore | Fundamentals of designing AR apps with machine learning models
37mARCore and indoor navigation, lessons learned
22mDeeper into ARKit with CoreML and Turi Create
29mFirebase MLKit — AI in Android Town!!