python我的Tflite模型在jupyter笔记本中运行良好。然而，当我试图在Android studio中运行它时，它向我显示了错误的输出

Question

因此，我根据患者输入的症状创建了用于疾病识别的自定义模型。它在python上运行得非常好，但是在我将其转换为tflite并在安卓 studio上运行之后，无论输入值如何，它都会给我错误的预测。 我已经在python上测试了tflite模型，它运行良好

以下代码显示了我如何将模型转换为tflite：-

    from tensorflow.keras.models import load_model
saved_model= load_model('Disease-predictor_ANN.h5')

    #Convert to tflite model
from tensorflow import lite 
converter= lite.TFLiteConverter.from_keras_model(saved_model)
tflite_model= converter.convert()

#Save the model
with open('dpredictor.tflite','wb') as f:
    f.write(tflite_model)

以下代码显示了我在安卓 studio中实现tflite模型的活动文件：-

package com.example.prototype_idoc;

import 安卓x.annotation.NonNull;
import 安卓x.appcompat.app.AppCompatActivity;


import 安卓.app.Activity;
import 安卓.content.Context;
import 安卓.content.pm.PackageManager;
import 安卓.content.res.AssetFileDescriptor;
import 安卓.content.res.AssetManager;
import 安卓.database.sqlite.SQLiteOpenHelper;
import 安卓.os.Bundle;
import 安卓.util.Log;
import 安卓.view.View;
import 安卓.widget.ArrayAdapter;
import 安卓.widget.Button;
import 安卓.widget.MultiAutoCompleteTextView;
import 安卓.widget.TextView;
import 安卓.widget.Toast;


import org.tensorflow.lite.DataType;
import org.tensorflow.lite.Interpreter;
import org.tensorflow.lite.support.tensorbuffer.TensorBuffer;
import org.tensorflow.lite.support.tensorbuffer.TensorBufferFloat;

import java.io.FileInputStream;
import java.io.IOException;
import java.lang.reflect.Array;
import java.nio.ByteBuffer;
import java.nio.ByteOrder;
import java.nio.FloatBuffer;
import java.nio.MappedByteBuffer;
import java.nio.channels.FileChannel;
import java.util.ArrayList;
import java.util.Arrays;

public class diseaseDiagnosis extends AppCompatActivity {

    private MultiAutoCompleteTextView multiAutoCompleteTextView;
    private Button button;
    private TextView textView;

    public diseaseDiagnosis() throws PackageManager.NameNotFoundException {
    }

    @Override
    protected void onCreate(Bundle savedInstanceState) {
        super.onCreate(savedInstanceState);
        setContentView(R.layout.activity_disease_diagnosis);

        MapdatabaseHelper mapdatabaseHelper= new MapdatabaseHelper(this);
        try {
            mapdatabaseHelper.createDatabase();
            mapdatabaseHelper.openDatabase();
        } catch (IOException e) {
            e.printStackTrace();
        }
        ArrayList<String> symptomList= (ArrayList<String>) mapdatabaseHelper.getEverySymptom(this);
        ArrayList<String> diseaseList= (ArrayList<String>) mapdatabaseHelper.getEveryDisease(this);
        String[] symptoms= symptomList.toArray(new String[symptomList.size()]);
        String[] diseases= diseaseList.toArray(new String[diseaseList.size()]);
        mapdatabaseHelper.close();
        multiAutoCompleteTextView= findViewById(R.id.multiAutoView);
        button= findViewById(R.id.checkDisease);
        textView= findViewById(R.id.diseaseName);

        ArrayAdapter<String> arrayAdapter= new ArrayAdapter<String>(this, 安卓.R.layout.simple_list_item_1,symptoms);
        multiAutoCompleteTextView.setAdapter(arrayAdapter);

        multiAutoCompleteTextView.setThreshold(1);
        multiAutoCompleteTextView.setTokenizer(new MultiAutoCompleteTextView.CommaTokenizer());


        button.setOnClickListener(new View.OnClickListener() {
            @Override
            public void onClick(View v) {

                float[] model_input= new float[132];
                Arrays.fill(model_input, new Float(0));
                String itemsSelected= multiAutoCompleteTextView.getText().toString();
                int startIndex=0;
                for(int i=0;i<itemsSelected.length();i++){
                    if(itemsSelected.charAt(i)==','){
                        String s=itemsSelected.substring(startIndex,i);
                        int currIndex=symptomList.indexOf(s);
                        Log.v("CURR INDEX:",s+ ":"+currIndex);
                        model_input[currIndex]= 1.0F;
                        startIndex=i+2;
                    }
                }

                Context context = null;
                try {
                    context = createPackageContext("com.example.prototype_idoc", 0);
                } catch (PackageManager.NameNotFoundException e) {
                    e.printStackTrace();
                }
                AssetManager assetManager = context.getAssets();
                Interpreter interpreter=null;
                try {
                   interpreter= new Interpreter(loadModelFile(assetManager,"dpredictor.tflite"));
                } catch (IOException e) {
                    e.printStackTrace();
                }

                ByteBuffer inputBuffer=ByteBuffer.allocateDirect(132*4).order(ByteOrder.nativeOrder());
                inputBuffer.rewind();
                for(int i=0;i<model_input.length;i++){
                    inputBuffer.putFloat(model_input[i]);
                }
                float[] outputArray= new float[41];
                TensorBuffer outputBuffer= TensorBuffer.createFixedSize(new int[]{1,41},DataType.FLOAT32);
                interpreter.run(inputBuffer,outputBuffer.getBuffer());
                outputArray= outputBuffer.getFloatArray();
//

                int modelOutputIndex= 0;
                float maxNum= outputArray[0];
                for(int i=0;i<outputArray.length;i++){
                    if(maxNum>outputArray[i]){
                        modelOutputIndex=i;
                        maxNum=outputArray[i];
                    }
                }
                Log.v("TAG:", "onClick: "+ outputArray[15] );
                Log.v("TAG:", "onClick: "+ outputArray[25] );
                textView.setText(diseases[modelOutputIndex]);
                textView.setVisibility(View.VISIBLE);
            }
        });



    }


    private MappedByteBuffer loadModelFile(AssetManager assetManager, String modelPath) throws IOException {
        AssetFileDescriptor fileDescriptor = assetManager.openFd(modelPath);
        FileInputStream inputStream = new FileInputStream(fileDescriptor.getFileDescriptor());
        FileChannel fileChannel = inputStream.getChannel();
        long startOffset = fileDescriptor.getStartOffset();
        long declaredLength = fileDescriptor.getDeclaredLength();
        return fileChannel.map(FileChannel.MapMode.READ_ONLY, startOffset, declaredLength);
    }