每日论文Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network
2017-03-10 16:15
771 查看
这篇文章用了深度残差网络同时用到GANs的概念构建的超分辨神经网络模型。主要贡献点：
1.构建由MSE优化的16blocks残差网络（SRResNet）。
2.SRGAN网络游GAN构建的网络。
3.MOS测试在3个数据库上SRGAN都不错的效果。
GAN的感知损失函数，除了GAN损失函数外还有基于内容的损失函数。内容的损失函数有两种，一种是MSE，还有一种是对VGG19的层卷积的结果求方差的损失函数。
网络结构如下图：
1.构建由MSE优化的16blocks残差网络（SRResNet）。
2.SRGAN网络游GAN构建的网络。
3.MOS测试在3个数据库上SRGAN都不错的效果。
GAN的感知损失函数，除了GAN损失函数外还有基于内容的损失函数。内容的损失函数有两种，一种是MSE，还有一种是对VGG19的层卷积的结果求方差的损失函数。
网络结构如下图：
内容来自用户分享和网络整理，不保证内容的准确性，如有侵权内容，可联系管理员处理 
相关文章推荐
- 论文阅读之：Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network
- Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network
- 《Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network》阅读笔记
- Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network
- Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network
- 《Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network》阅读笔记
- 《Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network》阅读笔记
- StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks 论文笔记
- “Single Image Super-resolution using Deformable Patches”
- 论文笔记之：Deep Generative Image Models using a Laplacian Pyramid of Adversarial Networks
- Image Super Resolution Using Conventional Neural Network (SRCNN)
- 论文笔记-深度估计(1)Depth Map Prediction from a Single Image using a Multi-Scale Deep Network
- Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural
- 深度学习论文笔记--Depth Map Prediction from a Single Image using a Multi-Scale Deep Network
- 【论文笔记】Depth Map Prediction from a Single Image using a Multi-Scale Deep Network
- Deeply-Recursive Convolutional Network for Image Super-Resolution
- 论文笔记——《Amortised MAP Inference for Image Super-Resolution》ICLR2017
- 每日论文Least Squares Generative Adversarial Networks
- 论文读书笔记-Using neural network to combine measures of word semantic similarity for image annotation
- ICCV2017论文“Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks” 阅读笔记