TensorFlowException：操作类型未注册＆＃39; NonMaxSuppressionV3＆＃39;在localhost上运行二进制文件

Question

我有一个Android应用程序，例如我对我的冻结模型进行了初始化。 似乎操作系统存在问题。我得到了这个例外：

java.lang.RuntimeException: Unable to start activity ComponentInfo{myapp/myapp.MainActivity}:
org.tensorflow.TensorFlowException: Op type not registered 'NonMaxSuppressionV3' in binary running on localhost. Make sure the Op and Kernel are registered in the binary running in this process.

at android.app.ActivityThread.performLaunchActivity(ActivityThread.java:2666)
    at android.app.ActivityThread.handleLaunchActivity(ActivityThread.java:2727)
    at android.app.ActivityThread.-wrap12(ActivityThread.java)
    at android.app.ActivityThread$H.handleMessage(ActivityThread.java:1478)
    at android.os.Handler.dispatchMessage(Handler.java:102)
    at android.os.Looper.loop(Looper.java:154)
    at android.app.ActivityThread.main(ActivityThread.java:6121)
    at java.lang.reflect.Method.invoke(Native Method)
    at com.android.internal.os.ZygoteInit$MethodAndArgsCaller.run(ZygoteInit.java:890)
    at com.android.internal.os.ZygoteInit.main(ZygoteInit.java:780)
 Caused by: org.tensorflow.TensorFlowException: Op type not registered 'NonMaxSuppressionV3' in binary running on localhost. Make sure the Op and Kernel are registered in the binary running in this process.
    at org.tensorflow.Graph.importGraphDef(Native Method)
    at org.tensorflow.Graph.importGraphDef(Graph.java:130)
    at org.tensorflow.Graph.importGraphDef(Graph.java:114)
    at org.tensorflow.contrib.android.TensorFlowInferenceInterface.loadGraph(TensorFlowInferenceInterface.java:559)
    at org.tensorflow.contrib.android.TensorFlowInferenceInterface.<init>(TensorFlowInferenceInterface.java:105)
    at myapp.ActivityInference.<init>(ActivityInference.java:22)
    at myapp.MainActivity.onCreate(MainActivity.java:21)
    at android.app.Activity.performCreate(Activity.java:6692)
    at android.app.Instrumentation.callActivityOnCreate(Instrumentation.java:1118)
    at android.app.ActivityThread.performLaunchActivity(ActivityThread.java:2619)
    at android.app.ActivityThread.handleLaunchActivity(ActivityThread.java:2727) 
    at android.app.ActivityThread.-wrap12(ActivityThread.java) 
    at android.app.ActivityThread$H.handleMessage(ActivityThread.java:1478) 
    at android.os.Handler.dispatchMessage(Handler.java:102) 
    at android.os.Looper.loop(Looper.java:154) 
    at android.app.ActivityThread.main(ActivityThread.java:6121) 
    at java.lang.reflect.Method.invoke(Native Method) 
    at com.android.internal.os.ZygoteInit$MethodAndArgsCaller.run(ZygoteInit.java:890)
    at com.android.internal.os.ZygoteInit.main(ZygoteInit.java:780)

我检查了tensorflow / core / kernels /中的BUILD文件，并定义了内核库：

tf_kernel_library( name = "non_max_suppression_op", prefix = "non_max_suppression_op", deps = IMAGE_DEPS, )

我是新手，所以我真的不知道如何解决它。 我正在研究预先训练好的冷冻模型。

这是我的Android代码：

package myapp;
import android.content.Context;
import android.content.res.AssetManager;

import org.tensorflow.contrib.android.TensorFlowInferenceInterface;


public class ActivityInference {

    private static ActivityInference activityInferenceInstance;
    private TensorFlowInferenceInterface inferenceInterface;
    private static AssetManager assetManager;
    private static final String MODEL_FILE = "file:///android_asset/frozen_inference_graph.pb";
    private static final String INPUT_NODE = "input";

    public ActivityInference(final Context context) {
        this.assetManager = context.getAssets();
        inferenceInterface = new TensorFlowInferenceInterface(assetManager, MODEL_FILE);

    }

    public static ActivityInference getInstance(final Context context){
        if (activityInferenceInstance == null)
        {
            activityInferenceInstance = new ActivityInference(context);
        }
        return activityInferenceInstance;
    }
}

以及Android的主要活动：

package myapp;

import android.support.v7.app.AppCompatActivity;
import android.os.Bundle;
import android.widget.TextView;


public class MainActivity extends AppCompatActivity {

    private ActivityInference activityInference;

    @Override
    protected void onCreate(Bundle savedInstanceState) {
        super.onCreate(savedInstanceState);
        setContentView(R.layout.activity_main);
        final TextView textView = (TextView) findViewById(R.id.textView);
        activityInference = new ActivityInference(getApplicationContext());
    }
}

我真的没有看到任何方法可以修复它，我想之后会有另外一个例外。 提前感谢任何人寻求解决方案。