-
Follow the Bazel steps for the TF Demo App:
-
Install Bazel and Android Prerequisites. It's easiest with Android Studio.
- You'll need at least SDK version 23.
- Make sure to install the latest version of Bazel. Some distributions ship with Bazel 0.5.4, which is too old.
- Bazel requires Android Build Tools
26.0.1
or higher. - Bazel is incompatible with NDK revisions 15 and above, with revision 16 being a compile-breaking change. Download an older version manually instead of using the SDK Manager.
- You also need to install the Android Support Repository, available
through Android Studio under
Android SDK Manager -> SDK Tools -> Android Support Repository
.
-
Edit your
WORKSPACE
to add SDK and NDK targets.NOTE: As long as you have the SDK and NDK installed, the
./configure
script will create these rules for you. Answer "Yes" when the script asks to automatically configure the./WORKSPACE
.- Make sure the
api_level
inWORKSPACE
is set to an SDK version that you have installed. - By default, Android Studio will install the SDK to
~/Android/Sdk
and the NDK to~/Android/Sdk/ndk-bundle
(but the NDK should be a manual download until Bazel supports NDK 16. See bullet points under (1)).
- Make sure the
-
Build the app with Bazel. The demo needs C++11:
bazel build -c opt --cxxopt='--std=c++11' \
//tensorflow/contrib/lite/java/demo/app/src/main:TfLiteCameraDemo
- Install the demo on a debug-enabled device:
adb install bazel-bin/tensorflow/contrib/lite/java/demo/app/src/main/TfLiteCameraDemo.apk