##########################################

Torch-7 for Android

########################################## Torch7 provides a Matlab-like environment for state-of-the-art machine learning algorithms. It is easy to use and provides a very efficient implementation, thanks to an easy and fast scripting language (Lua) and a underlying C implementation.

Modified to be compiled and used with Android

Features

• Loading of lua packages from the apk directly.
• This is done by writing a custom package.loader Reference: http://www.lua.org/manual/5.1/manual.html#pdf-package.loaders The loader is in torchandroid.cpp as loader_android
• torchandroid.h and torchandroid.cpp give lots of helper functions to make life easier
  • Print function overriden to redirect to logcat (only handles strings for now)
  • Function to get apk assets as bytes (very useful)

Requirements

Android NDK and Android SDK

Samples

• A sample project has been provided in android-demo
• android-demo/jni/torchdemo.cpp is a simple use-case
• android-demo/assets/main.lua is the file that is run
• Vinayak Ghokale from e-lab Purdue (https://github.com/e-lab) contributed a face detector demo, which showcases a fuller use-case.
• That's in the facedetector_e-lab folder. I made some changes to it to load assets etc. from apk as opposed to the sdcard, but it remains untouched otherwise.

Building Torch

1. open build.sh and modify ANDROID_NDK_TOOLCHAIN_ROOT to your android ndk path.
2. run build script $ sh build.sh

You can use torch in your android apps. The relevant directories are

• include - include directories
• lib - static libs cross-compiled for armeabi-v7a
• share - lua files

Building Example

1. Build Torch atleast once using the steps above.
2. [Optional] Connect your android phone in debugging mode, to automatically install the apk.
3. Change directory into android-demo folder.
4. Run build script. $ sh build.sh
5. Run the app TorchDemo on your phone.