基于树莓pi的tensorflow lite量化ssd目标检测

在使用C++的tensorflow lite的树莓pi上，对象检测没有按预期工作。我的代码编译并运行，但输出似乎从未得到正确填充。我可能缺少任何依赖项或访问结果错误吗？

我遵循了以下教程：https://medium.com/tensorflow/training-and-serving-a-realtime-mobile-object-detector-in-30-minutes-with-cloud-tpus-b78971cf1193

并且具有来自以下的detect.tflite模型：https://storage.googleapis.com/download.tensorflow.org/models/tflite/pets_ssd_mobilenet_v1_0.75_quant_2018_06_29.zip

我已经为树莓pi编译了tensorflow lite和opencv，并修改了minimal.cc以读取图像并执行如下推理：

/* Copyright 2018 The TensorFlow Authors. All Rights Reserved.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
==============================================================================*/
#include <cstdio>
#include "tensorflow/contrib/lite/interpreter.h"
#include "tensorflow/contrib/lite/kernels/register.h"
#include "tensorflow/contrib/lite/model.h"
#include "tensorflow/contrib/lite/optional_debug_tools.h"
#include <iostream>
#include "opencv2/core.hpp"
#include "opencv2/imgcodecs.hpp"
// This is an example that is minimal to read a model
// from disk and perform inference.
//
// Usage: detect <tflite model> <image filepath>
using namespace tflite;
#define TFLITE_MINIMAL_CHECK(x)                              
if (!(x)) {                                                
fprintf(stderr, "Error at %s:%dn", __FILE__, __LINE__); 
exit(1);                                                 
}
int main(int argc, char* argv[]) {
if(argc != 3) {
fprintf(stderr, "minimal <tflite model filepath> <image filepath>n");
return 1;
} else {
fprintf(stdout, "Reading model from %sn", argv[1]);
fprintf(stdout, "Reading image from %sn", argv[2]);
}
const char* model_filename = argv[1];
const char* image_filename = argv[2];
// Load model
std::unique_ptr<tflite::FlatBufferModel> model =
tflite::FlatBufferModel::BuildFromFile(model_filename);
TFLITE_MINIMAL_CHECK(model != nullptr);
// Build the interpreter
tflite::ops::builtin::BuiltinOpResolver resolver;
InterpreterBuilder builder(*model.get(), resolver);
std::unique_ptr<Interpreter> interpreter;
builder(&interpreter);
TFLITE_MINIMAL_CHECK(interpreter != nullptr);
// Allocate tensor buffers.
TFLITE_MINIMAL_CHECK(interpreter->AllocateTensors() == kTfLiteOk);
printf("=== Pre-invoke Interpreter State ===n");
tflite::PrintInterpreterState(interpreter.get());
// Fill input buffers
// TODO(user): Insert code to fill input tensors
cv::Mat img = cv::imread(image_filename);
//std::cout << "before: " << interpreter->typed_input_tensor<uchar>(0) << std::endl;
//std::cout << "image: " << img.data << std::endl;
memcpy(interpreter->typed_input_tensor<uchar>(0), img.data, img.total() * img.elemSize());
//std::cout << "after: " << interpreter->typed_input_tensor<uchar>(0) << std::endl;
// Run inference
TFLITE_MINIMAL_CHECK(interpreter->Invoke() == kTfLiteOk);
printf("nn=== Post-invoke Interpreter State ===n");
tflite::PrintInterpreterState(interpreter.get());
// Read output buffers
// TODO(user): Insert getting data out code.
cv::Mat results0(10, 4, CV_8U);
cv::Mat results1(1, 10, CV_8U);
cv::Mat results2(1, 10, CV_8U);
cv::Mat results3(1, 1, CV_8U);
results0.data = interpreter->typed_output_tensor<uchar>(0);
results1.data = interpreter->typed_output_tensor<uchar>(1);
results2.data = interpreter->typed_output_tensor<uchar>(2);
results3.data = interpreter->typed_output_tensor<uchar>(3);
std::cout << "results 0: " << results0 << std::endl;
std::cout << "results 1: " << results1 << std::endl;
std::cout << "results 2: " << results2 << std::endl;
std::cout << "results 3: " << results3 << std::endl;
return 0;
}

我看到的结果是：

results 0: []               
results 1: []               
results 2: []               
results 3: []