superResNet.h

/*
 * Copyright (c) 2019, NVIDIA CORPORATION. All rights reserved.
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the "Software"),
 * to deal in the Software without restriction, including without limitation
 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
 * and/or sell copies of the Software, and to permit persons to whom the
 * Software is furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
 * DEALINGS IN THE SOFTWARE.
 */

#ifndef __SUPERRESNET_NET_H__
#define __SUPERRESNET_NET_H__


#include "tensorNet.h"

/**
 * @note superResNet is only supported with TensorRT 5.0 and newer,
 * as it uses ONNX models and requires ONNX import support in TensorRT.
 */
#if NV_TENSORRT_MAJOR >= 5
#  define HAS_SUPERRES_NET
#endif
/**
 * Name of default input blob for segmentation model.
 * @ingroup deepVision
 */
#define RESNET_DEFAULT_INPUT "input_0"

/**
 * Name of default output blob for segmentation model.
 * @ingroup deepVision
 */
#define RESNET_DEFAULT_OUTPUT "output_0"
/**
 * Super Resolution Network
 */
class superResNet : public tensorNet {
 public:
  /**
	 * Load super resolution network
	 */
  enum NetworkType {
    FCN_ResNet18_Cityscapes_512x256,
    FCN_ResNet18_Cityscapes_1024x512,
    FCN_ResNet18_Cityscapes_2048x1024,
    FCN_ResNet18_DeepScene_576x320,
    FCN_ResNet18_DeepScene_864x480,
    FCN_ResNet18_MHP_512x320,
    FCN_ResNet18_MHP_640x360,
    FCN_ResNet18_Pascal_VOC_320x320,
    FCN_ResNet18_Pascal_VOC_512x320,
    FCN_ResNet18_SUN_RGBD_512x400,
    FCN_ResNet18_SUN_RGBD_640x512
  };

  static superResNet* Create();
  static superResNet* Create(const char* model_path,
                            const char* input = RESNET_DEFAULT_INPUT,
                            const char* output = RESNET_DEFAULT_OUTPUT,
                            uint32_t maxBatchSize = 2);
  static superResNet* Create(NetworkType networkType,
                            uint32_t maxBatchSize = 2);
  
                

  /**
	 * Destroy
	 */
  ~superResNet();

  /**
	 * Upscale a 4-channel RGBA image.
	 */
  bool UpscaleRGBA(float* input, uint32_t inputWidth, uint32_t inputHeight,
                   float* output, uint32_t outputWidth, uint32_t outputHeight,
                   float maxPixelValue = 255.0f);

  /**
	 * Upscale a 4-channel RGBA image.
	 */
  bool UpscaleRGBA(float* input, float* output, float maxPixelValue = 255.0f);

  /**
	 * Retrieve the width of the input image, in pixels.
	 */
  inline uint32_t GetInputWidth() const { return mWidth; }

  /**
	 * Retrieve the height of the input image, in pixels.
	 */
  inline uint32_t GetInputHeight() const { return mHeight; }

  /**
	 * Retrieve the width of the output image, in pixels.
	 */
  inline uint32_t GetOutputWidth() const { return DIMS_W(mOutputs[0].dims); }

  /**
	 * Retrieve the height of the output image, in pixels.
	 */
  inline uint32_t GetOutputHeight() const { return DIMS_H(mOutputs[0].dims); }

  /**
	 * Retrieve the scale factor between the input and output.
	 */
  inline double GetScaleFactor() const {
    return (double)(double(GetOutputWidth()) / double(GetInputWidth()));
  }

 protected:
  superResNet();
};


#endif