步骤：

一、在app的build.gradle中加入相关依赖。其内容自行百度。

二、加载模型：

 /**
     * 此方法实现对 .tflite 文件的序列化
     * @param model
     * @return
     * @throws IOException
     */
    private MappedByteBuffer loadModelFile(String model) throws IOException {
        AssetFileDescriptor fileDescriptor = getApplicationContext().getAssets().openFd(model + ".tflite");
        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);
    }


    /**
     * 对序列化后的结果进行加载
     * @param model
     */
    private void load_model(String model) {
        try {
            tflite = new Interpreter(loadModelFile(model));
            Toast.makeText(MainActivity.this, model + " model load success", Toast.LENGTH_SHORT).show();
            Log.d(TAG, model + " model load success");
            tflite.setNumThreads(4);
            load_result = true;
        } catch (IOException e) {
            Toast.makeText(MainActivity.this, model + " model load fail", Toast.LENGTH_SHORT).show();
            Log.d(TAG, model + " model load fail");
            load_result = false;
            e.printStackTrace();
        }
    }

三、运行结果：

Interpreter tflite = null;
...
//调用以上两个函数实现tflite的初始化
...

tflite.run(inputData,labelProbArray);
tflite.close(); //释放资源

参考https://www.tensorflow.org/lite/guide/inference