Run ONNX models on your Android phone using the new NNAPI !
Android 8.1 introduces Neural Networks API (NNAPI). It's very exciting to run a model in the "native" way supported by Android System. :)
DNNLibrary is a wrapper of NNAPI ("DNNLibrary" is for "daquexian's NNAPI library). It lets you easily make the use of the new NNAPI introduced in Android 8.1. You can convert your onnx model into daq
and run the model directly.
For the Android app example, please check out dnnlibrary-example.
Telegram Group: link, QQ Group (Chinese): 948989771, answer: 哈哈哈哈
This screenshot is ResNet-18
This screenshot is LeNet
Clone this repo and submodules:
git clone --recursive https://github.com/daquexian/DNNLibrary
Please make sure the Android System on your phone is 8.1+, or you may want to use an 8.1+ emulator.
Android 8.1 introduces NNAPI. However, NNAPI is not friendly to normal Android developers. It is not designed to be used by normal developers directly. So I wrapped it into a library.
With DNNLibrary it's extremely easy to deploy your ONNX model on Android 8.1+ phone. For example, following is the Java code to deploy the MobileNet v2 in your app (please check out daq-example for detail):
ModelBuilder modelBuilder = new ModelBuilder();
Model model = modelBuilder.readFile(getAssets(), "mobilenetv2.daq")
.setOutput("mobilenetv20_output_pred_fwd"); // The output name is from the onnx model
.compile(ModelBuilder.PREFERENCE_FAST_SINGLE_ANSWER);
float[] result = model.predict(inputData);
Only five lines! And the daq
model file is got from the pretrained onnx model using onnx2daq
.
After cloning step listed in Preparation section, run
mkdir build
cd build
cmake ..
cmake --build .
Now onnx2daq
is in tools/onnx2daq
directory. The following command is to convert onnx model to daq model.
./tools/onnx2daq/onnx2daq <onnx model> <output filename>
For example, if you have a model named "mobilenetv2.onnx" in your current directory,
./tools/onnx2daq/onnx2daq mobilenetv2.onnx mobilenetv2.daq
Welcome! It has been published on jcenter.
Add
implementation 'me.daquexian:dnnlibrary:0.5.3'
(for Gradle 3.0+),
or
compile 'me.daquexian:dnnlibrary:0.5.3'
(for Gradle lower than 3.0)
in your app's build.gradle
's dependencies
section.
We use CMake as the build system. So you can build it as most C++ projects, the only difference is that you need Android NDK, r17b or higher NDK is necessary :
mkdir build && cd build
cmake -DCMAKE_SYSTEM_NAME=Android -DCMAKE_ANDROID_NDK=<path of android ndk> -DCMAKE_ANDROID_ARCH_ABI=arm64-v8a -DCMAKE_ANDROID_NDK_TOOLCHAIN_VERSION=clang -DCMAKE_ANDROID_STL_TYPE=c++_static -DCMAKE_SYSTEM_VERSION=<Android API level, 27 or 28> ..
cmake --build .
then you will get binary files.
Yes, but its support for NNAPI is far from perfect. For example, dilated convolution (which is widely used in segmentation) are not supported (https://github.com/tensorflow/tensorflow/blob/master/tensorflow/contrib/lite/nnapi_delegate.cc#L458).
What's more, only the models got from TensorFlow can easily get converted to TF Lite model. Since NNAPI is independent of any frameworks, we support ONNX, a framework-independent model format.
_ | TF Lite | DNNLibrary |
---|---|---|
Supported Model Format | TensorFlow | ONNX |
Dilated Convolution | ❌ | ✔️ |
Ease of Use | ❌ (Bazel build system, not friendly to Android developers) |
✔️ |
Quantization | ✔️ | ❌ (WIP, plan to base on this) |
However we are also far from maturity comparing to TF Lite. At least we are an another choice if you want to enjoy the power of NNAPI :)
Google Pixel, Android 9.0
model | time |
---|---|
MobileNet v2 | 132.95ms |
SqueezeNet v1.1 | 80.80ms |
More benchmark is welcome!
The old DNNLibrary supports caffe model by dnntools, however, it is not supported directly now, the models generated by dnntools
are not usable, too. Please use a convert tool like MMdnn to convert the caffe model to the ONNX model, then convert it to daq
using onnx2daq
.