This library helps with getting started with TensorFlow Lite on Android. Inspired by TensorFlow Lite Android image classification example.

This is an experimental library and subject to change. It's written entirely in Kotlin and powered by TensorFlow Lite.

The library provides helper class for Image Classification at the minimum usage. In future releases, the library may provide helpers for Object Detection and Smart Reply.

Gradle

implementation 'com.neno0o.tflitehelper:imageclassification:0.0.1'

Maven

<dependency>
  <groupId>com.neno0o.tflitehelper</groupId>
  <artifactId>imageclassification</artifactId>
  <version>0.0.1</version>
  <type>pom</type>
</dependency>

Add TensorFlow Lite library

implementation "org.tensorflow:tensorflow-lite:latest_version"

To get started with Image Classification, get instance by providing your model and label paths in the asset folder to create factory method

private lateinit var imageClassification: ImageClassification

imageClassification = ImageClassification.create(
                classifierModel = ClassifierModel.QUANTIZED,
                assetManager = assets,
                modelPath = MODEL_PATH,
                labelPath = LABEL_PATH
            )

To classify (recognize) an image, call classifyImage with bitmap..

val results = imageClassification.classifyImage(bitmap)

The results are list of Recognizable.

The sample app uses Coroutines while initializing ImageClassification and when calling classifyImage to not block the main thread.

val results = async { imageClassification.classifyImage(bitmap) }

withContext(Dispatchers.Main) {
    object_name.text = results.await().toString()
}

Check out the sample app for more details.

ImageClassification uses default input size 224, you should override this value depending on your model. Also, you can override the number of results that will be returned, and the default confidence threshold.

imageClassification = ImageClassification.create(
                classifierModel = ClassifierModel.QUANTIZED,
                assetManager = assets,
                modelPath = MODEL_PATH,
                labelPath = LABEL_PATH,
                inputSize = YOUR_MODEL_INPUT_SIZE,
                numberOfResults = 10,
                confidenceThreshold = .5f
            )

To configure the Interpreter via the Interpreter.Options, create instance of Interpreter.Options alongside its configurations such as the number of threads that are used to configure the Interpreter, and use it in create method.

private lateinit var imageClassification: ImageClassification

val interpreterOptions = Interpreter.Options()
interpreterOptions.setNumThreads(4)
interpreterOptions.setUseNNAPI(true) // If Android device supports NNAPI

imageClassification = ImageClassification.create(
                classifierModel = ClassifierModel.QUANTIZED,
                assetManager = assets,
                modelPath = MODEL_PATH,
                labelPath = LABEL_PATH,
                interpreterOptions = interpreterOptions
            )

Copyright 2019 Ahmed Gamal. All Rights Reserved.

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.