openvino/inference-engine/samples/validation_app/Processor.hpp

95 lines
3.1 KiB
C++

/*
// Copyright (c) 2018 Intel Corporation
//
// 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.
*/
#pragma once
#include <iostream>
#include <limits>
#include <string>
#include <memory>
#include <samples/common.hpp>
#include "inference_engine.hpp"
#include "csv_dumper.hpp"
#include "image_decoder.hpp"
#include "console_progress.hpp"
using namespace std;
#define OUTPUT_FLOATING(val) std::fixed << std::setprecision(2) << val
class Processor {
public:
struct InferenceMetrics {
int nRuns = 0;
double minDuration = std::numeric_limits<double>::max();
double maxDuration = 0;
double totalTime = 0;
virtual ~InferenceMetrics() { } // Type has to be polymorphic
};
protected:
std::string modelFileName;
std::string targetDevice;
std::string imagesPath;
int batch;
InferenceEngine::InferRequest inferRequest;
InferenceEngine::InputsDataMap inputInfo;
InferenceEngine::OutputsDataMap outInfo;
InferenceEngine::CNNNetReader networkReader;
InferenceEngine::SizeVector inputDims;
InferenceEngine::SizeVector outputDims;
double loadDuration;
PreprocessingOptions preprocessingOptions;
CsvDumper& dumper;
InferencePlugin plugin;
std::string approach;
double Infer(ConsoleProgress& progress, int filesWatched, InferenceMetrics& im);
public:
Processor(const std::string& flags_m, const std::string& flags_d, const std::string& flags_i, int flags_b,
InferenceEngine::InferencePlugin plugin, CsvDumper& dumper, const std::string& approach, PreprocessingOptions preprocessingOptions);
virtual shared_ptr<InferenceMetrics> Process() = 0;
virtual void Report(const InferenceMetrics& im) {
double averageTime = im.totalTime / im.nRuns;
slog::info << "Inference report:\n";
slog::info << "\tNetwork load time: " << loadDuration << "ms" << "\n";
slog::info << "\tModel: " << modelFileName << "\n";
slog::info << "\tModel Precision: " << networkReader.getNetwork().getPrecision().name() << "\n";
slog::info << "\tBatch size: " << batch << "\n";
slog::info << "\tValidation dataset: " << imagesPath << "\n";
slog::info << "\tValidation approach: " << approach;
slog::info << slog::endl;
if (im.nRuns > 0) {
slog::info << "Average infer time (ms): " << averageTime << " (" << OUTPUT_FLOATING(1000.0 / (averageTime / batch))
<< " images per second with batch size = " << batch << ")" << slog::endl;
} else {
slog::warn << "No images processed" << slog::endl;
}
}
virtual ~Processor() {}
};