Keynote! The opportunity for Android devs with AI and ML
summary
Android developers get concrete routes to ship ML models on-device using TensorFlow -> TensorFlow Lite, Model Maker and MediaPipe Tasks, with a preview no-code GUI coming later; a built app Project Guideline guides blind runners via on-phone camera and works in the field, diabetic retinopathy models hit ~95–98% accuracy and a retina model predicts gender ~97% from scans, and workflows target bitmap/tensor conversion, live camera pipelines and privacy by running models entirely on device.
AI-written summary · about
chapters (13)
- 0:00Welcome
- 1:03Economic plateaus and tech platforms
- 3:52AI as next dev platform
- 5:55Coding vs ML paradigms
- 9:02How ML models learn
- 12:10AI-enabled apps
- 13:33TensorFlow on Android
- 16:23Future of Android dev
- 21:39TensorFlow Lite and Model Maker
- 27:34Data pipelines for ML
- 30:19Custom models and no-code
- 33:08Resources for Android AI
- 36:20Q&A
related talks
39mMachine Learning on Android Demystified
43mWhat’s new in Android machine learning
39mAI/ML on mobile/android with TensorFlow
47mMyths and Reality of Android Machine Learning
32mAdvanced on-device ML made easy
47mMachine Learning on Android Demystified
13mWhat's new in Android machine learning
36mMaking intelligent Android apps