PyTorch Mobile Kit

PyTorch Mobile Kit is a starter kit app that does Machine Learning on edge from camera output, photos, and videos.

Demo

Watch on YouTube.

---

New (Dec 2021) Build AI-powered mobile apps in minutes with PyTorch Live (Beta).

PyTorch Live brings together PyTorch and React Native, making it easier to build cross-platform mobile apps.

Live give you an easy-to-use library of tools for bringing on-device AI demos to life -- easily integrate vision and language models into your apps.

Project Structure

The code for the Android project is in this PyTorchMobileKit directory.

Machine Learning Models

Currently, the Android app are using pre-trained Computer Vision model, which is packaged in TorchVision. These models are optimized for offline and low-latency inference on mobile devices:

• MobileNet v2
• ResNet-18
• ResNeXt-50 32x4d
• SqueezeNet 1.1

Software Requirements

• Python 3.6+
• PyTorch 1.3+

---

Get Started

Expand Get Started on Android

Tutorial with a Basic Example

BasicApp is a simple image classification application that demonstrates how to use PyTorch Android API.

This application runs TorchScript serialized TorchVision pretrained Resnet-18 model on static image which is packaged inside the app as Android asset.

1. Model Preparation

Let’s start with model preparation. If you are familiar with PyTorch, you probably should already know how to train and save your model. In case you don’t, we are going to use a pre-trained image classification model (Resnet18), which is packaged in TorchVision.

To install it, run the command below:

pip install torchvision

To serialize the model you can use Python scripts in the model directory:

import torch
import torchvision

model = torchvision.models.resnet18(pretrained=True)
model.eval()
input = torch.rand(1, 3, 224, 224)
traced_script_module = torch.jit.trace(model, input)
traced_script_module.save("../BasicApp/app/src/main/assets/resnet18.pt")

If everything works well, we should have our model - resnet18.pt generated in the assets directory of Android application.

That will be packaged inside Android application as asset and can be used on the device.

More details about TorchScript you can find in tutorials on pytorch.org.

2. Cloning from GitHub

git clone https://github.com/cedrickchee/pytorch-mobile-kit.git
cd BasicApp

If Android SDK and Android NDK are already installed you can install this application to the connected android device or emulator with:

./gradlew installDebug

We recommend you to open this project in Android Studio 3.5.1+ (At the moment PyTorch Android and demo applications use Android gradle plugin of version 3.5.0, which is supported only by Android Studio version 3.5.1 and higher), in that case you will be able to install Android NDK and Android SDK using Android Studio UI.

3. Gradle Dependencies

Pytorch Android is added to the BasicApp as gradle dependencies in build.gradle:

repositories {
    jcenter()
}

dependencies {
    implementation 'org.pytorch:pytorch_android:1.3.0'
    implementation 'org.pytorch:pytorch_android_torchvision:1.3.0'
}

where org.pytorch:pytorch_android is the main dependency with PyTorch Android API, including libtorch native library for all 4 Android abis (armeabi-v7a, arm64-v8a, x86, x86_64). In this doc, you can find how to rebuild it from source only for specific list of Android abis.

org.pytorch:pytorch_android_torchvision - additional library with utility functions for converting android.media.Image and android.graphics.Bitmap to tensors.

4. Loading TorchScript Module

Module module = Module.load(assetFilePath(this, "model.pt"));

org.pytorch.Module represents torch::jit::script::Module that can be loaded with load method specifying file path to the serialized to file model.

5. Preparing Input

Tensor inputTensor = TensorImageUtils.bitmapToFloat32Tensor(bitmap,
    TensorImageUtils.TORCHVISION_NORM_MEAN_RGB, TensorImageUtils.TORCHVISION_NORM_STD_RGB);

org.pytorch.torchvision.TensorImageUtils is part of org.pytorch:pytorch_android_torchvision library.

The TensorImageUtils#bitmapToFloat32Tensor method creates tensors in the torchvision format using android.graphics.Bitmap as a source.

All pre-trained models expect input images normalized in the same way, i.e. mini-batches of 3-channel RGB images of shape (3 x H x W), where H and W are expected to be at least 224. The images have to be loaded in to a range of [0, 1] and then normalized using mean = [0.485, 0.456, 0.406] and std = [0.229, 0.224, 0.225]

inputTensor's shape is 1x3xHxW, where H and W are bitmap height and width appropriately.

6. Run Inference

Tensor outputTensor = module.forward(IValue.from(inputTensor)).toTensor();
float[] scores = outputTensor.getDataAsFloatArray();

org.pytorch.Module.forward method runs loaded module's forward method and gets result as org.pytorch.Tensor outputTensor with shape 1x1000.

7. Processing Results

Its content is retrieved using org.pytorch.Tensor.getDataAsFloatArray() method that returns Java array of floats with scores for every ImageNet class.

After that we just find index with maximum score and retrieve predicted class name from Constants.IMAGENET_CLASSES array that contains all ImageNet classes.

float maxScore = -Float.MAX_VALUE;
int maxScoreIdx = -1;
for (int i = 0; i < scores.length; i++) {
  if (scores[i] > maxScore) {
    maxScore = scores[i];
    maxScoreIdx = i;
  }
}
String className = Constants.IMAGENET_CLASSES[maxScoreIdx];

Background

Previous attempts:

• PyTorch Lite
• PyTorch on Android

License

This repository contains a variety of content; some developed by Cedric Chee, and some from third-parties. The third-party content is distributed under the license provided by those parties.

I am providing code and resources in this repository to you under an open source license. Because this is my personal repository, the license you receive to my code and resources is from me and not my employer.

The content developed by Cedric Chee is distributed under the following license:

Code

The code in this repository is released under the MIT license.