Now smile (also in 3D)! Exploring AR, ML and Camera related APIs on Android

on YouTube ↗44m

" summary

Live demos show an AR app that anchors persistent labels in 3D and combines local and cloud ML: CameraX for simple photo/video, ARCore + ML Kit + TensorFlow Lite for on-device object detection, and Cloud Vision API for freshness. Code from Google samples runs and mostly works; offline TF Lite models miss some objects while cloud calls add latency but return richer, up‑to‑date results. Recommendation: use CameraX for basic capture, TF Lite for low‑latency specialized models, and cloud APIs when model currency matters.

AI-written summary · about

" chapters (9)

1. 0:00Introduction
2. 6:45Android camera APIs
3. 13:07Camera beyond photos
4. 16:18ARCore on Android
5. 19:10ML camera APIs
6. 27:57Combining AR, ML and camera APIs
7. 37:35Live demos
8. 41:05Summary
9. 42:52Q&A

" related talks

• 39mNow smile (also in 3D)! Exploring AR, ML and Camera related APIs on AndroidWalmyr Carvalho · Droidcon London 2023
• 41mFusing AI with AR: Custom MarkersStephen Wylie · Windcity DevFest 2019
• 21mLive Coding: Adding Machine Learning to Android Things in 10 MinsHoi Lam · droidcon Berlin 2018
• 42mHello ARCoreGiovanni Laquidara · Codemotion Milan 2017
• 31mRecreating That AR App with 2019 TechnologyKenichi Takahashi · DroidKaigi 2019
• 34mIt’s an ML full world! MLKit for Android DevsBritt Barak · DevFest Ukraine 2018
• 37mExploring Augmented Reality with ARCoreChristophe Jollivet, Jérémy Voisin · DevFest Toulouse 2019
• 41mMaking Android Apps with IntelligenceMargaret Maynard Reid · Devoxx US 2017