VRT: A Video Restoration Transformer | ✓ Link | 7.628 | 0.758 | 0.929 | 0.902 | 0.722 | 31.669 | 2.778 | VRT | 2022-01-28 |
BasicVSR: The Search for Essential Components in Video Super-Resolution and Beyond | ✓ Link | 7.186 | 0.75 | 0.934 | 0.9 | 0.709 | 31.443 | 2.128 | BasicVSR | 2020-12-03 |
Recurrent Back-Projection Network for Video Super-Resolution | ✓ Link | 7.068 | 0.746 | 0.74 | 0.899 | 0.629 | 31.407 | 0.043 | RBPN | 2019-03-25 |
Deep Blind Video Super-resolution | ✓ Link | 6.947 | 0.737 | 0.921 | 0.894 | 0.629 | 31.071 | 0.241 | DBVSR | 2020-03-10 |
iSeeBetter: Spatio-Temporal Video Super Resolution using Recurrent-Generative Back-Projection Networks | ✓ Link | 6.809 | 0.748 | 0.741 | 0.896 | 0.629 | 31.104 | 0.045 | iSeeBetter | 2020-06-21 |
Local-Global Fusion Network for Video Super-Resolution | ✓ Link | 6.505 | 0.74 | 0.903 | 0.898 | 0.629 | 31.291 | 0.667 | LGFN | 2020-09-22 |
DynaVSR: Dynamic Adaptive Blind Video Super-Resolution | ✓ Link | 6.136 | 0.709 | 0.884 | 0.865 | 0.557 | 28.377 | 0.177 | DynaVSR-R | 2020-11-09 |
Temporal Modulation Network for Controllable Space-Time Video Super-Resolution | ✓ Link | 6 | 0.712 | 0.931 | 0.885 | 0.549 | 30.364 | 1.136 | TMNet | 2021-04-21 |
COMISR: Compression-Informed Video Super-Resolution | ✓ Link | 5.637 | 0.654 | 0.879 | 0.84 | 0.619 | 26.708 | 1.613 | COMISR | 2021-05-04 |
Video Super-Resolution with Recurrent Structure-Detail Network | ✓ Link | 5.566 | 0.667 | 0.819 | 0.826 | 0.619 | 25.321 | 1.961 | RSDN | 2020-08-02 |
Video Super-resolution with Temporal Group Attention | ✓ Link | 5.529 | 0.669 | 0.859 | 0.831 | 0.549 | 25.786 | 0.706 | TGA | 2020-07-21 |
TDAN: Temporally Deformable Alignment Network for Video Super-Resolution | ✓ Link | 5.454 | 0.706 | 0.721 | 0.883 | 0.609 | 30.244 | 0.493 | TDAN | 2018-12-07 |
Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data | ✓ Link | 5.392 | 0.663 | 0.895 | 0.774 | 0 | 24.441 | 1.01 | Real-ESRGAN | 2021-07-22 |
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks | ✓ Link | 5.353 | 0.735 | 0.948 | 0.808 | 0 | 27.33 | 1.004 | ESRGAN | 2018-09-01 |
Revisiting Temporal Modeling for Video Super-resolution | ✓ Link | 5.35 | 0.627 | 0.842 | 0.79 | 0.557 | 24.252 | 2.567 | RRN-10L | 2020-08-13 |
Deep Video Super-Resolution Network Using Dynamic Upsampling Filters Without Explicit Motion Compensation | ✓ Link | 5.324 | 0.645 | 0.87 | 0.83 | 0.549 | 25.852 | 0.418 | DUF-28L | 2018-06-01 |
Real-World Super-Resolution via Kernel Estimation and Noise Injection | ✓ Link | 5.286 | 0.69 | 0.911 | 0.767 | 0 | 25.989 | 0.352 | RealSR | 2020-06-19 |
Deep Video Super-Resolution Network Using Dynamic Upsampling Filters Without Explicit Motion Compensation | ✓ Link | 5.124 | 0.641 | 0.868 | 0.828 | 0.549 | 24.606 | 0.605 | DUF-16L | 2018-06-01 |
Deformable 3D Convolution for Video Super-Resolution | ✓ Link | 5.066 | 0.674 | 0.915 | 0.876 | 0.549 | 29.703 | 0.041 | D3Dnet | 2020-04-06 |
Revisiting Temporal Modeling for Video Super-resolution | ✓ Link | 5.02 | 0.617 | 0.856 | 0.789 | 0.549 | 23.786 | 2.74 | RRN-5L | 2020-08-13 |
Deep Video Super-Resolution using HR Optical Flow Estimation | ✓ Link | 4.863 | 0.647 | 0.895 | 0.831 | 0.557 | 25.986 | 0.699 | SOF-VSR-BD | 2020-01-06 |
Deep Video Super-Resolution using HR Optical Flow Estimation | ✓ Link | 4.805 | 0.66 | 0.904 | 0.872 | 0.557 | 29.381 | 0.571 | SOF-VSR-BI | 2020-01-06 |
SwinIR: Image Restoration Using Swin Transformer | ✓ Link | 4.799 | 0.618 | 0.895 | 0.782 | 0 | 25.12 | 0.407 | SwinIR | 2021-08-23 |
DynaVSR: Dynamic Adaptive Blind Video Super-Resolution | ✓ Link | 4.359 | 0.643 | 0.859 | 0.864 | 0.549 | 29.011 | 0.15 | DynaVSR-V | 2020-11-09 |
Hierarchical Conditional Flow: A Unified Framework for Image Super-Resolution and Image Rescaling | ✓ Link | 4.262 | 0.713 | 0.923 | 0.791 | 0 | 26.067 | 0.066 | HCFlow | 2021-08-11 |
Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data | ✓ Link | 3.697 | 0.598 | 0.871 | 0.824 | 0 | 27.195 | 1.019 | Real-ESRnet | 2021-07-22 |
Learning a Single Convolutional Super-Resolution Network for Multiple Degradations | ✓ Link | 3.468 | 0.594 | 0.877 | 0.834 | 0 | 27.672 | 5.882 | SRMD | 2017-12-17 |
[]() | | 3.308 | 0.6 | 0.861 | 0.838 | 0 | 27.716 | 1.282 | waifu2x-cunet | |
[]() | | 2.739 | 0.57 | 0.856 | 0.819 | 0 | 27.176 | 3.125 | waifu2x-anime | |
Towards Real-World Blind Face Restoration with Generative Facial Prior | ✓ Link | 2.686 | 0.538 | 0.793 | 0.745 | 0 | 24.195 | 1.562 | GFPGAN | 2021-01-11 |
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network | ✓ Link | 2.099 | 0.521 | 0.765 | 0.811 | 0 | 26.714 | 3.333 | ESPCN | 2016-09-16 |
Blind Face Restoration via Deep Multi-scale Component Dictionaries | ✓ Link | 0.277 | 0.339 | 0.623 | 0.759 | 0 | 24.832 | 0.909 | DFDnet | 2020-08-02 |