How to use tensorflow pre-trained model in Android

本文主要介绍在Retrain a tensorflow model based on Inception v3基础上，将新训练的tensorflow模型移植到Android中。

阅读本文前，请确保已经知晓Retrain a tensorflow model based on Inception v3中内容。

准备工作

$ cd ~
$ git clone https://github.com/googlecodelabs/tensorflow-for-poets-2
$ cd ~/tensorflow-for-poets-2
$ cp -r ~/tf_files .
$ ls -lah tf_files/ # 查看内容应该包含模型文件retrained_graph.pb和retrained_labels.txt

• 下载安装Android Studio

优化模型文件

Optimize for inference

$ cd ~/tensorflow-for-poets-2
$ python -m tensorflow.python.tools.optimize_for_inference \
  --input=tf_files/retrained_graph.pb \
  --output=tf_files/optimized_graph.pb \
  --input_names="Cast" \
  --output_names="final_result"

Quantize the network weights

$ cd ~/tensorflow-for-poets-2
$ python -m scripts.quantize_graph \
  --input=tf_files/optimized_graph.pb \
  --output=tf_files/rounded_graph.pb \
  --output_node_names=final_result \
  --mode=weights_rounded

评估量化网络前后，模型的整体识别准确率。

$ cd ~/tensorflow-for-poets-2
$ python -m scripts.evaluate  tf_files/optimized_graph.pb
$ python -m scripts.evaluate  tf_files/rounded_graph.pb

You should see less than a 1% change in the model accuracy. With the reduced dataset you may see no change in accuracy at all.

设置Android App

准备优化后的模型文件

$ cd ~/tensorflow-for-poets-2
$ cp tf_files/rounded_graph.pb tf_files/retrained_labels.txt android/assets/ 

引入TensorFlow AAR文件

在project的build.gradle文件中添加以下内容：

allprojects {
    repositories {
        jcenter()
    }
}

dependencies {
    compile 'org.tensorflow:tensorflow-android:+'
}

This will tell Gradle to use the latest version of the TensorFlow AAR that has been released to https://bintray.com/google/tensorflow/tensorflow-android. You may replace the + with an explicit version label if you wish to use a specific release of TensorFlow in your app.

插入手机

1. 启动手机的开发者模式；
2. 启动 开发者选项，USB调试 和 USB安装；
3. 通过数据线链接手机；

运行

1. 启动Android Studio
2. 点击Open an existing Android Studio project
3. 选择tensorflow-for-poets-2/android文件夹

等待编译完成后，启动运行，并在手机开机（未锁屏）状态下，点击安装。

在启动手机上的app后，需要点击同意TensorFlowDemo照相和记录视频，同意TensorFlowDemo操作手机上的图片,媒体和文件。

参考

1. TensorFlow For Poets 2
2. Android TensorFlow support
3. TensorFlow Android Camera Demo