Update app to work with Caffe2 from PyTorch master

* Android app is working with my own `resnet18_init_net_v1.pb` and `resnet18_predict_net_v1.pb` Protobuf files

Author: cedrickchee

.idea/caches/build_file_checksums.ser

@@ -6,15 +6,96 @@ PyTorch on Android is a project to demo how to use [PyTorch](https://pytorch.org
 
 _The source code for the demo in this repo was originally based on [AICamera repo](https://github.com/bwasti/AICamera) unchanged._
 
-**Project Status**: Early release. Still in heavy development. What this means is, things might be moved around quickly and things will break.
+**Project Status**:
+
+- Early release. Still in heavy development. What this means is, things might be moved around quickly and things will break.
+- 2018-12-31:
+  - PyTorch core maintainers have updated AICamera example to work with latest PyTorch master. Once that PR is merged into PyTorch master, you can use the README below to get a working Android app, including changing the Protobuf with your own `init.pb` / `predict.pb` files.
+  - [Android OSS fixes PR](https://github.com/pytorch/pytorch/pull/15509).
+- 2019-01-01:
+  - I have tested the Android OSS fixes with my own `resnet18_init_net_v1.pb` and `resnet18_predict_net_v1.pb` Protobuf files and the Android app is working fine.
 
 ## Goal
 
 Make it easier to ship and test your neural network model in PyTorch on mobile devices.
 
 ## Documentation
 
-We created a developer guide on [how to ship a convolutional neural network on Android with PyTorch and Android Studio](https://github.com/cedrickchee/data-science-notebooks/blob/master/notebooks/deep_learning/fastai_mobile/README.md). Check it out!
+### Updating the AICamera Android app to work with Caffe2 from PyTorch/master
+
+This is an example for using Caffe2 on Android.
+
+1. git clone PyTorch source and switch to remote branch, `android_oss_fixes`:
+
+```sh
+$ git clone --recursive https://github.com/pytorch/pytorch.git
+$ cd pytorch
+# We are using this PyTorch master commit 39381964bbb2686cf84c78aed638985455f6d3fe (Dec 23 02:08:33 2018 -0500, "fix build_android for newest NDK, and make install work")
+$ git checkout --track origin/android_oss_fixes
+
+M third_party/QNNPACK
+M third_party/cpuinfo
+M third_party/fbgemm
+M third_party/ideep
+Branch 'android_oss_fixes' set up to track remote branch 'android_oss_fixes' from 'origin'.
+Switched to a new branch 'android_oss_fixes'
+```
+
+2. You'll need to download the [Android NDK (latest r18 release)](https://developer.android.com/ndk/downloads/) if you have not.
+
+3. First, we will compile Caffe2 Android libraries like libcaffe2, libqnnpack for arm-v7a ABI and then for x86.
+
+Set build environment:
+- PyTorch folder is at `$PYTORCH_ROOT`
+- This repository folder is at `$AICAMERA_ROOT`
+- Android NDK folder is at `$ANDROID_NDK`
+
+```sh
+# make sure $PYTORCH_ROOT, $AICAMERA_ROOT and $ANDROID_NDK are set
+export ANDROID_NDK=~/android/sdk/ndk-bundle/
+export PYTORCH_ROOT=~/dev/gh/pytorch/
+export AICAMERA_ROOT=~/dev/android/android-studio-projects/aicamera/
+
+pushd $PYTORCH_ROOT
+```
+
+Then, do the following:
+
+Build Caffe2 android libs and copy them over into AICamera app folder
+
+```sh
+./scripts/build_android.sh
+```
+If you encountered build errors related to "quantized/int8_*.cc", get the patch from this "[Update QNNPACK](https://github.com/pytorch/pytorch/pull/15561/files)" PR. Patch these files in your local copies.
+
+You should see "Install configuration: "Release"" output in your terminal if the process completed successfully.
+
+```sh
+mv build_android build_android_arm
+
+# copy headers
+cp -r install/include/* $AICAMERA_ROOT/app/src/main/cpp/
+
+# copy arm libs
+rm -rf $AICAMERA_ROOT/app/src/main/jniLibs/armeabi-v7a/
+mkdir $AICAMERA_ROOT/app/src/main/jniLibs/armeabi-v7a
+cp -r build_android_arm/lib/lib* $AICAMERA_ROOT/app/src/main/jniLibs/armeabi-v7a/
+
+
+./scripts/build_android.sh -DANDROID_ABI=x86
+mv build_android build_android_x86
+
+# copy x86 libs
+rm -rf $AICAMERA_ROOT/app/src/main/jniLibs/x86/
+mkdir $AICAMERA_ROOT/app/src/main/jniLibs/x86
+cp -r build_android_x86/lib/lib* $AICAMERA_ROOT/app/src/main/jniLibs/x86/
+```
+
+3. Build the AICamera app using the `Build -> Make Project` menu option in Android Studio
+
+### Developer Guide
+
+We created a developer guide on [how to ship a convolutional neural network (Resnet18 and SqueezeNet) on Android with PyTorch and Android Studio](https://github.com/cedrickchee/data-science-notebooks/blob/master/notebooks/deep_learning/fastai_mobile/README.md). Check it out!
 
 ### Step by Step Guide
 
@@ -26,7 +107,7 @@ We'll walk you through every step, from problem all the way to building and depl
 - [x] Test SqueezeNet v1.1 model with your own video stream from camera
 - [x] Live (real-time) detection
 - [ ] Performance optimization
-- [ ] Test build in models (MobileNetV2, ShuffleNet) with your own video stream from camera
+- [x] Test build in models (MobileNetV2, ShuffleNet, Resnet18) with your own video stream from camera
 - [ ] Download your own model on the fly and test it
 - [ ] Manage models locally on your Android device
 - [ ] Memory consumption and time elapse data
@@ -36,7 +117,7 @@ We'll walk you through every step, from problem all the way to building and depl
 - [ ] New classifier for insect species dataset using pre-trained ImageNet weights (transfer learning)
 - [ ] Upload pre-trained models
 - [ ] Bug fixes
-- [ ] Fix intermittent crashes
+- [x] Fix intermittent crashes
 - [ ] React Native native module
 
 ## Android Project
@@ -49,17 +130,18 @@ You can download the Android project source code by running this command:
 
 ### Dependencies
 
-- Android Studio: 3.0 and above
+- Android Studio: 3.2.1 and above
 - Android SDK
-- Android NDK
+- Android NDK r18 and above
 - CMake Android SKD component
-- Gradle 3.x
+- Gradle 4.6 and above
 
 ## Tests
 
 | Device             | Network       |  FPS (^)  |
 | ------------------ | ------------- | --------- |
 | Google Nexus 6P    | SqueezeNet    |  3.0      |
+| Google Nexus 6P    | Resnet18      |  0.6      |
 | Samsung Note 8     | SqueezeNet    |  _TBD_    |
 | Galaxy Note 3      | SqueezeNet    |  4.0      |
 | Samsung Galaxy S7  | SqueezeNet    |  5.8      |