TensorFlow for Mobile Developers

Transcript

1. TensorFlow for Mobile Developers

2. Machine Learning / AI

3. Buzzwords classification Artificial intelligence Machine Learning The science of making

   machines smarts Building machines that can learn Neural Networks One of the many different algorithms in Machine Learning

4. Buzzwords classification

5. Neural Network

6. Machine Learning

7. Machine Learning

8. Machine Learning 12% of all responses on mobile

9. Machine Learning

10. Machine Learning

11. Machine Learning

12. Cloud vs. Mobile Less Traffic & Faster Response Motion Sensors

13. TensorFlow OpenSource library for Machine Learning tensorflow.org Today most popular

    ML framework

14. TensorFlow You can train: - Mac / Windows - GPU

    Server - GPU/TPU on Cloud Prediction: - Android - iOS - Raspberry

15. TensorFlow CIFAR10

16. TensorFlow ecosystem

17. From Training to App (@yufengg)

18. Data Gathering?

19. From Training to App (@yufengg)

20. From Training to App (@yufengg)

21. Convolutional Neural Networks (@yufengg)

22. Convolutional Neural Networks (@yufengg)

23. Convolutional Neural Networks (@yufengg)

24. Convolutional Neural Networks (@yufengg)

25. Convolutional Neural Networks (@yufengg)

26. Convolutional Neural Networks (@yufengg)

27. Convolutional Neural Networks (@yufengg)

28. Convolutional Neural Networks (@yufengg)

29. From Training to App (@yufengg)

30. Optimizing model for Mobile (@yufengg)

31. From Training to App (@yufengg) Other inception versions (inception v1

    quantised is 7 MB)

32. From Training to App (@yufengg) TensorFlow increases APK in 12

    MB

33. From Training to App (@yufengg)

34. From Training to App (@yufengg)

35. TensorFlow community 1000+ contributors 22.000+ commits 18.000+ repositories with name

    “TensorFlow”

36. TensorFlow community

37. Building with Android Studio Uses Bazel to build Set Bazel

    binary location in /tensorflow/examples/ android/build.gradle Add project in tensorflor/examples/android folder to Android Studio

38. Building with Android Studio Installing Bazel first: Download from https://github.com/bazelbuild/bazel/releases

    Install instructions in https://docs.bazel.build/versions/master/install.html

39. Building with Android Studio Download Android SDK (23) Download Android

    NDK (12b)
 
 (recommend it with Android SDK Manager)

40. Building with Android Studio WORKSPACE FILE

41. Building with Android Studio Run Bazel bazel build -c opt

    //tensorflow/examples/ android:tensorflow_demo

42. Building with Android Studio Install APK adb install -r bazel-bin/tensorflow/examples/android/

    tensorflow_demo.apk

43. Building with Android Studio OR… TensorFlow AAR from JCenter YUV

    -> RGB less efficient Object tracking not available

44. Android Samples TF Classify 
 TF Detect
 
 TF Stylize

45. TF Classify Uses Google Inception (v3) to label images Model

    easy to swap No “person label” Volume button up for statistics

46. TF Classify

47. TF Classify

48. TF Detect

49. TF Detect - Draw bounding boxes around people - Useful

    to count objects - No training yet

50. TF Style

51. TF Style - Real time style transfer algorithm - Pick/mix

    different styles - Can train Magenta models

52. TF Style

53. TensorFlow Written in C++ Android uses Kotlin/Java?
 
 HowTo? Android

    inference Library https://github.com/tensorflow/tensorflow/tree/master/tensorflow/contrib/android

54. Building for iOS Install requirements: - Xcode 8 - Command

    line tools (xcode-select install) - brew install automake - brew install lib tool - tensorflow/contrib/makefile_build_all_ios.sh (takes 20 minutes)

55. iOS Examples: simple

56. iOS Examples: camera

57. iOS Examples: camera - Run inception each frame - Models

    can be replaced

58. iOS Examples: benchmark

59. Raspberry Pi

60. Resources TensorFlow: https://www.tensorflow.org/ Magenta: https://magenta.tensorflow.org/welcome-to-magenta TensorFlow for Poets: https://codelabs.developers.google.com/codelabs/ tensorflow-for-poets/index.html#0