A simple Android example that demonstrates skin cancer classification using tflite and gpu-delegates

The build.gradle is configured to use TensorFlow Lite's nightly build.

Here we are using the gpu-delegates 0.0.0, which was first experimental version. There is a newer version as well but with huge changes. As of now, we will work with the older version. You need to do follow these steps:

• Step 1. Clone the TensorFlow source code and open it in Android Studio

    git clone https://github.com/tensorflow/tensorflow
  

• Step 2. Edit app/build.gradle to use the nightly GPU AAR. Add the tensorflow-lite-gpu package alongside the existing tensorflow-lite package in the existing dependencies block. Please make sure you have correct version of Android SDK

       dependencies {
           ...
           implementation 'org.tensorflow:tensorflow-lite:0.0.0-nightly'
           implementation 'org.tensorflow:tensorflow-lite-gpu:0.0.0-nightly'
       }
  

• Step 3. Build and run Run → Run ‘app’. When you run the application you will see a button for enabling the GPU. You can toggle between CPU and GPU. You will see the performance difference between both.