TensorFlow for Mobile Developers

Transcript

1. TensorFlow for Mobile Developers Enrique López Mañas Google Developer Expert
2. Ego Slide • Freelance Dev • Google Developer Expert • @eenriquelopez
3. Machine Learning / AI
4. Machine Learning / AI
5. Machine Learning / AI
6. 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
7. Buzzwords classification
8. Neural Network
9. Machine Learning
10. Machine Learning
11. Machine Learning 12% of all responses on mobile
12. Machine Learning
13. Machine Learning
14. Machine Learning
15. Cloud vs. Mobile Less Traffic & Faster Response Motion Sensors
16. TensorFlow OpenSource library for Machine Learning tensorflow.org Today most popular ML framework
17. TensorFlow You can train: - Mac / Windows - GPU Server - GPU/TPU on Cloud Prediction: - Android - iOS - Raspberry
18. TensorFlow CIFAR10
19. TensorFlow ecosystem
20. From Training to App (@yufengg)
21. Data Gathering?
22. From Training to App (@yufengg)
23. From Training to App (@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. Convolutional Neural Networks (@yufengg)
30. Convolutional Neural Networks (@yufengg)
31. Convolutional Neural Networks (@yufengg)
32. From Training to App (@yufengg)
33. Optimizing model for Mobile (@yufengg)
34. From Training to App (@yufengg) Other inception versions (inception v1 quantised is 7 MB)
35. From Training to App (@yufengg) TensorFlow increases APK in 12 MB
36. From Training to App (@yufengg)
37. From Training to App (@yufengg)
38. TensorFlow community 1000+ contributors 22.000+ commits 18.000+ repositories with name “TensorFlow”
39. TensorFlow community
40. 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
41. 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
42. Building with Android Studio Download Android SDK (23) Download Android NDK (12b)
 
 (recommend it with Android SDK Manager)
43. Building with Android Studio WORKSPACE FILE
44. Building with Android Studio Run Bazel bazel build -c opt //tensorflow/examples/ android:tensorflow_demo
45. Building with Android Studio Install APK adb install -r bazel-bin/tensorflow/examples/android/ tensorflow_demo.apk
46. Building with Android Studio OR… TensorFlow AAR from JCenter YUV -> RGB less efficient Object tracking not available
47. Android Samples TF Classify 
 TF Detect
 
 TF Stylize
48. TF Classify Uses Google Inception (v3) to label images Model easy to swap No “person label” Volume button up for statistics
49. TF Classify
50. TF Classify
51. TF Detect
52. TF Detect - Draw bounding boxes around people - Useful to count objects - No training yet
53. TF Style
54. TF Style - Real time style transfer algorithm - Pick/mix different styles - Can train Magenta models
55. TF Style
56. TensorFlow Written in C++ Android uses Kotlin/Java?
 
 HowTo? Android inference Library https://github.com/tensorflow/tensorflow/tree/master/tensorflow/contrib/android
57. 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)
58. iOS Examples: simple
59. iOS Examples: camera
60. iOS Examples: camera - Run inception each frame - Models can be replaced
61. iOS Examples: benchmark
62. Raspberry Pi
63. 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
64. Feedback http://bit.ly/Mobius2017