Superpower Your Android apps with ML: Android 11

In this session, my major aim would be to provide an overview of the different tools one could use to power their Android apps with Machine Learning and also discuss the new additions for Machine Learning in Android 11 specifically the Model Binding Plugin and ML Kit. I would first go on to explain the standard procedure of using pre-trained models with MLKit. I would show how we could take the idea of MLKit forward and use pre-trained models from TensorFlow Hub to run right in the app, which would provide support to build high-quality machine learning apps based on models contributed from the community. I would then show how we could use custom TFLite models in Android apps, I would also talk about TensorFlow Model Maker and ML Model binding plugin in Android Studio through which I plan to show how easy it is to now use custom TF Lite models in Android apps. With Android 11 the NN API now supports Asymmetric integer weights making model sizes and inferences even smaller opening up even larger opportunities for edge ML.
If time persists, I would also show demos about the above topics.

Transcript

1. rishit.tech Superpower Your Android apps with ML Rishit Dagli @rishit_dagli
2. rishit.tech $whoami • 11 Grade Student • TEDx and Ted-Ed Speaker • ♡ Hackathons and competitions • ♡ Research • My coordinates - https://www.rishit.tech rishit_dagli Rishit-dagli
3. Acknowledgements • Sayak Paul (ML GDE) • Khanh LeViet (Google) • Hoi Lam (Google) rishit.tech
4. rishit.tech Ideal Audience • Mobile Devs looking for ways to build smarter apps • Mobile Devs looking for ways to integrate ML in their existing apps easily
5. rishit.tech Why care about ML in Android?
6. rishit.tech Why care about on-device ML in Android?
7. Created by Rishit Dagli for his talk at GDG Ahmedabad This slide is skipped while presenting
8. Created by Rishit Dagli for his talk at GDG Ahmedabad This slide is skipped while presenting
9. rishit.tech
10. rishit.tech
11. rishit.tech
12. rishit.tech
13. rishit.tech Why should you care? • Power consumption
14. rishit.tech Why should you care? • Power consumption • Inference time
15. rishit.tech Why should you care? • Power consumption • Inference time • Network availability
16. rishit.tech Why should you care? • Power consumption • Inference time • Network availability • Privacy
17. rishit.tech ML Model Binding Plugin
18. rishit.tech ML Model Binding Plugin What’s new for on-device ML in Android? Easier to use Enable Hardware acceleration Faster Development
19. rishit.tech Importing a TF Lite Model
20. rishit.tech Importing a TF Lite Model
21. rishit.tech Importing a TF Lite Model
22. rishit.tech Using the TF Lite Model
23. rishit.tech Creating an instance of model
24. rishit.tech Processing images
25. rishit.tech Passing in data
26. rishit.tech Passing in data
27. rishit.tech Passing in data
28. rishit.tech Passing in data
29. rishit.tech Adding Labels rishit.tech
30. rishit.tech We are done! ☺ rishit.tech
31. rishit.tech GPU acceleration
32. rishit.tech A new ML Kit
33. rishit.tech A new ML Kit What does the latest ML Kit focus on? On-Device ML Better customizability Generic use cases
34. rishit.tech Face detection Barcode scanning Image labeling Smart Reply Language Identification Vision Natural Language Object detection and tracking On-device Translation Text recognition Digital Ink Recognition Pose Detection N EW N EW
35. rishit.tech Setup the model
36. rishit.tech Customize the model
37. rishit.tech
38. rishit.tech TF Lite Model Maker • High performance • Super easy to use • High customization too!
39. rishit.tech TF Hub tfhub.dev
40. rishit.tech TF Hub
41. rishit.tech Demos! bit.ly/dc-apac
42. rishit.tech Q & A rishit_dagli Rishit-dagli
43. rishit.tech Slides bit.ly/dc-apac-slides
44. rishit.tech Thank you! rishit_dagli Rishit-dagli