Super-resolution¶
some cross-project repo:
Martin Krasser/Keras-super-resolution: SRGAN, EDSR, WDSR
open-mmlab/MMSR: SRResNet, SRGAN, ESRGAN, EDVR, etc.
SRCNN¶
Image Super-Resolution Using Deep Convolutional Networks (TPAMI 2014) by CUHK
Project - include Matlab and Caffe code
first upscale it to the desired size using bicubic interpolation
FSRCNN¶
Accelerating the Super-Resolution Convolutional Neural Network (ECCV 2016) by CUHK
use deconvolution to upscale instead of pre-processing
VDSR¶
Accurate Image Super-Resolution Using Very Deep Convolutional Networks (CVPR 2016)
start to use residual, one long skip-connection
Same as SRCNN, it take upscaled LR image as imput
SRGAN¶
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network (CVPR 2017) by Twitter
Tensorflow 2.0
SRResNet + long skip-connection, diverge from MSE
novel perceptual loss using high-level feature maps of VGG network
EDSR¶
Enhanced Deep Residual Networks for Single Image Super-Resolution (CVPR 2017) remove batch normalization of SRResNet→ save resource spend resource on convolution instead
MDSR¶
multiscale architecture that reconstructs various scales of high-resolution images in a single model
WDSR¶
Wide Activation for Efficient and Accurate Image Super-Resolution (CVPR Workshop 2018)
Official pyTorch (based on EDSR) | tensorflow
- increase channel before ReLU
- linear low-rank convolution stack and large channel
SFTGAN¶
Recovering Realistic Texture in Image Super-resolution by Deep Spatial Feature Transform (CVPR 2018) - SenseTime + CUHK
pyTorch + Lua | STF Layer
- Spatial Feature Transform: learns a mapping function \(M\) that outputs a modulation parameter pair (γ, β) based on some prior condition Ψ (Segmentation probability maps)
- generate affine transformation parameters for spatial-wise (pixel-wise) feature modulation
RDN¶
Residual Dense Network for Image Super-Resolution (CVPR 2018)
- Residual + Dense block
- local feature fusion + global feature fusion -> contiguous memory mechanism used in video debluring/BIN
ESRGAN¶
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks (ECCV 2018) - CUHK
- RRDB (Residual in Residual Dense Blcok)
- enhance adversarial loss and perceptual loss
- relativistic GAN
- improve the perceptual loss by using the features before activation