App detect and Classifier using Yolov3 on Android

## TOOL - OpenCV 3.4.3 - Android Studio - Tensorflow 1.13.1 (Should using correctly version) - DW2TF-1.1 or DW2TF-1.2 (Don't using DW2TF-1.0) - Yolov3 Tiny ## MAIN ### 1. Convert Yolov3-tiny model of darknet to darkflow - clone DW2TF-1.2 (in last release) from here [DW2TF-1.2](https://github.com/jinyu121/DW2TF/releases/tag/v1.2) Problem with DW2TF-1.0 (Yolov3, Yolov3-tiny,... be affected). What is problem? [Problem](jinyu121/DW2TF#30) - Download (or train) Yolov3 tiny model and config file in darknet (.cfg and .weight). - Launch DW2TF conversion as mentioned on the github ```python3 python3 main.py \ --cfg 'data/yolov3-tiny.cfg' \ --weights 'data/yolov3-tiny.weights' \ --output 'data/' \ --prefix 'yolov3-tiny/' \ --gpu 0 ``` - Conver Darkflow to tensorflow for android. Launch freeze_graph to have a single bp graph file: ```python3 freeze_graph \ --input_graph yolov3-tiny.pb \ --input_checkpoint yolov3-tiny.ckpt \ --input_binary=true \ --output_graph=ultimate_yolov3-tiny.bp \ --output_node_names="yolov3-tiny/convolutional10/BiasAdd, yolov3-tiny/convolutional13/BiasAdd" ``` - Note: For older version of Yolo you can use darkflow tool [Order](https://github.com/thtrieu/darkflow),Load: ```python3 ./flow --model ../data/yolov2-tiny.cfg --load ../data/yolov2-tiny.weights --savepb ``` You can download [Yolo](https://pjreddie.com/darknet/yolo/) at offical page

2. Java code in android

Add package nesscessary

• build.gradle(Module: App)

.....
repositories {
    maven {
        url 'https://google.bintray.com/tensorflow'
    }
}
dependencies {
    implementation fileTree(dir: "libs", include: ["*.jar"])
    implementation 'androidx.appcompat:appcompat:1.2.0'
    implementation 'androidx.constraintlayout:constraintlayout:2.0.4'
    testImplementation 'junit:junit:4.12'
    androidTestImplementation 'androidx.test.ext:junit:1.1.2'
    androidTestImplementation 'androidx.test.espresso:espresso-core:3.3.0'
    //implementation 'com.android.support:appcompat-v7:27.1.1'
    implementation 'com.android.support.constraint:constraint-layout:1.1.3'
    testImplementation 'junit:junit:4.12'
    androidTestImplementation 'com.android.support.test:runner:1.0.2'
    androidTestImplementation 'com.android.support.test.espresso:espresso-core:3.0.2'
//    androidTestImplementation 'androidx.test.espresso:espresso-core:3.1.1'
//    androidTestImplementation 'androidx.test:runner:1.1.1'
    androidTestImplementation 'androidx.test:rules:1.1.1'
    implementation 'org.tensorflow:tensorflow-android:1.13.1'
    implementation 'org.apache.commons:commons-lang3:3.0'
    implementation project(':openCVLibrary343')
}

• settings.gradle( Project setting)

include ':app'
rootProject.name = "TinyYolo"
include ':openCVLibrary343'

Open some manefests necessary

 <uses-permission android:name="android.permission.WRITE_EXTERNAL_STORAGE"/>
    <uses-permission android:name="android.permission.READ_EXTERNAL_STORAGE"/>
    <uses-permission android:name="android.permission.CAMERA"/>
    <uses-feature android:name="android.hardware.camera" android:required="false"/>
    <uses-feature android:name="android.hardware.camera.autofocus" android:required="false"/>
    <uses-feature android:name="android.hardware.camera.front" android:required="false"/>
    <uses-feature android:name="android.hardware.camera.front.autofocus" android:required="false"/>

RecognizeImage using tensorflow in Java

• Training model

- Main processing

- Main Algorithm

- Android Async Task

### 3. Result - App identifies and classifies correctly, but lacks