Tasks SotA Datasets Papers Methods Submit About

Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable Benchmarks All SotA Datasets Papers Methods

Community

Submit Results About

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Models/bicubic

bicubic

Reported on 91 benchmarks across 13 tasks · 4 papers · 11 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision43 results

3D Human Pose EstimationonVid4 - 4x upscaling
MOVIE· 2016-11-16
9.31
SOTA
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
VideoonVid4 - 4x upscaling
MOVIE· 2016-11-16
9.31
SOTA
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
Pose EstimationonVid4 - 4x upscaling
MOVIE· 2016-11-16
9.31
SOTA
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
Video Super-ResolutiononVid4 - 4x upscaling
MOVIE· 2016-11-16
9.31
SOTA
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
3D Object Super-ResolutiononVid4 - 4x upscaling
MOVIE· 2016-11-16
9.31
SOTA
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
Image Super-ResolutiononManga109 - 4x upscaling
PSNR· 2018-09-01
24.89
best: 33.19 (HMA†)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
Image Super-ResolutiononManga109 - 4x upscaling
SSIM· 2018-09-01
0.7866
best: 0.9366 (Hi-IR-L)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
Image Super-ResolutiononUrban100 - 4x upscaling
PSNR· 2018-09-01
23.14
best: 28.72 (Hi-IR-L)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
Image Super-ResolutiononUrban100 - 4x upscaling
SSIM· 2018-09-01
0.6577
best: 0.9481 (SPSR)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
3D Object Super-ResolutiononManga109 - 4x upscaling
PSNR· 2018-09-01
24.89
best: 33.19 (HMA†)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
3D Object Super-ResolutiononManga109 - 4x upscaling
SSIM· 2018-09-01
0.7866
best: 0.9366 (Hi-IR-L)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
3D Object Super-ResolutiononUrban100 - 4x upscaling
PSNR· 2018-09-01
23.14
best: 28.72 (Hi-IR-L)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
3D Object Super-ResolutiononUrban100 - 4x upscaling
SSIM· 2018-09-01
0.6577
best: 0.9481 (SPSR)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
3D Human Pose EstimationonVid4 - 4x upscaling
PSNR· 2016-11-16
23.82
best: 31.36 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
3D Human Pose EstimationonVid4 - 4x upscaling
SSIM· 2016-11-16
0.6548
best: 0.933 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
VideoonVid4 - 4x upscaling
PSNR· 2016-11-16
23.82
best: 31.36 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
VideoonVid4 - 4x upscaling
SSIM· 2016-11-16
0.6548
best: 0.933 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
Pose EstimationonVid4 - 4x upscaling
PSNR· 2016-11-16
23.82
best: 31.36 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
Pose EstimationonVid4 - 4x upscaling
SSIM· 2016-11-16
0.6548
best: 0.933 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
Video Super-ResolutiononVid4 - 4x upscaling
PSNR· 2016-11-16
23.82
best: 31.36 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
Video Super-ResolutiononVid4 - 4x upscaling
SSIM· 2016-11-16
0.6548
best: 0.933 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
3D Object Super-ResolutiononVid4 - 4x upscaling
PSNR· 2016-11-16
23.82
best: 31.36 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
3D Object Super-ResolutiononVid4 - 4x upscaling
SSIM· 2016-11-16
0.6548
best: 0.933 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
3D Human Pose EstimationonUltra Video Group HD - 4x upscaling
Average PSNR· 2016-09-16
36.2
best: 48.23 (RAMS (ours))
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
3D Human Pose EstimationonXiph HD - 4x upscaling
Average PSNR· 2016-09-16
30.3
best: 31.67 (ESPCN)
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
VideoonUltra Video Group HD - 4x upscaling
Average PSNR· 2016-09-16
36.2
best: 48.23 (RAMS (ours))
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
VideoonXiph HD - 4x upscaling
Average PSNR· 2016-09-16
30.3
best: 31.67 (ESPCN)
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
Pose EstimationonUltra Video Group HD - 4x upscaling
Average PSNR· 2016-09-16
36.2
best: 48.23 (RAMS (ours))
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
Pose EstimationonXiph HD - 4x upscaling
Average PSNR· 2016-09-16
30.3
best: 31.67 (ESPCN)
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
Video Super-ResolutiononUltra Video Group HD - 4x upscaling
Average PSNR· 2016-09-16
36.2
best: 48.23 (RAMS (ours))
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
Video Super-ResolutiononXiph HD - 4x upscaling
Average PSNR· 2016-09-16
30.3
best: 31.67 (ESPCN)
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
3D Object Super-ResolutiononUltra Video Group HD - 4x upscaling
Average PSNR· 2016-09-16
36.2
best: 48.23 (RAMS (ours))
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
3D Object Super-ResolutiononXiph HD - 4x upscaling
Average PSNR· 2016-09-16
30.3
best: 31.67 (ESPCN)
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
Image Super-ResolutiononSet14 - 4x upscaling
MOS· 2016-09-15
1.8
best: 3.72 (SRGAN)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
Image Super-ResolutiononSet14 - 4x upscaling
SSIM· 2016-09-15
0.7486
best: 0.894 (Edge-informed SR)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
Image Super-ResolutiononBSD100 - 4x upscaling
MOS· 2016-09-15
1.47
best: 3.56 (SRGAN)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
Image Super-ResolutiononBSD100 - 4x upscaling
PSNR· 2016-09-15
25.94
best: 28.16 (DRCT-L)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
Image Super-ResolutiononBSD100 - 4x upscaling
SSIM· 2016-09-15
0.6935
best: 0.851 (Edge-informed SR)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
3D Object Super-ResolutiononSet14 - 4x upscaling
MOS· 2016-09-15
1.8
best: 3.72 (SRGAN)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
3D Object Super-ResolutiononSet14 - 4x upscaling
SSIM· 2016-09-15
0.7486
best: 0.894 (Edge-informed SR)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
3D Object Super-ResolutiononBSD100 - 4x upscaling
MOS· 2016-09-15
1.47
best: 3.56 (SRGAN)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
3D Object Super-ResolutiononBSD100 - 4x upscaling
PSNR· 2016-09-15
25.94
best: 28.16 (DRCT-L)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
3D Object Super-ResolutiononBSD100 - 4x upscaling
SSIM· 2016-09-15
0.6935
best: 0.851 (Edge-informed SR)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802

Graphs14 results

Super-ResolutiononVid4 - 4x upscaling
MOVIE· 2016-11-16
9.31
SOTA
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
Super-ResolutiononManga109 - 4x upscaling
PSNR· 2018-09-01
24.89
best: 33.19 (HMA†)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
Super-ResolutiononManga109 - 4x upscaling
SSIM· 2018-09-01
0.7866
best: 0.9366 (Hi-IR-L)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
Super-ResolutiononUrban100 - 4x upscaling
PSNR· 2018-09-01
23.14
best: 28.72 (Hi-IR-L)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
Super-ResolutiononUrban100 - 4x upscaling
SSIM· 2018-09-01
0.6577
best: 0.9481 (SPSR)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
Super-ResolutiononVid4 - 4x upscaling
PSNR· 2016-11-16
23.82
best: 31.36 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
Super-ResolutiononVid4 - 4x upscaling
SSIM· 2016-11-16
0.6548
best: 0.933 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
Super-ResolutiononUltra Video Group HD - 4x upscaling
Average PSNR· 2016-09-16
36.2
best: 48.23 (RAMS (ours))
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
Super-ResolutiononXiph HD - 4x upscaling
Average PSNR· 2016-09-16
30.3
best: 31.67 (ESPCN)
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
Super-ResolutiononSet14 - 4x upscaling
MOS· 2016-09-15
1.8
best: 3.72 (SRGAN)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
Super-ResolutiononSet14 - 4x upscaling
SSIM· 2016-09-15
0.7486
best: 0.894 (Edge-informed SR)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
Super-ResolutiononBSD100 - 4x upscaling
MOS· 2016-09-15
1.47
best: 3.56 (SRGAN)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
Super-ResolutiononBSD100 - 4x upscaling
PSNR· 2016-09-15
25.94
best: 28.16 (DRCT-L)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
Super-ResolutiononBSD100 - 4x upscaling
SSIM· 2016-09-15
0.6935
best: 0.851 (Edge-informed SR)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802

Methodology14 results

3DonVid4 - 4x upscaling
MOVIE· 2016-11-16
9.31
SOTA
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
16konManga109 - 4x upscaling
PSNR· 2018-09-01
24.89
best: 33.19 (HMA†)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
16konManga109 - 4x upscaling
SSIM· 2018-09-01
0.7866
best: 0.9366 (Hi-IR-L)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
16konUrban100 - 4x upscaling
PSNR· 2018-09-01
23.14
best: 28.72 (Hi-IR-L)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
16konUrban100 - 4x upscaling
SSIM· 2018-09-01
0.6577
best: 0.9481 (SPSR)
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks arXiv:1809.00219
3DonVid4 - 4x upscaling
PSNR· 2016-11-16
23.82
best: 31.36 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
3DonVid4 - 4x upscaling
SSIM· 2016-11-16
0.6548
best: 0.933 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
3DonUltra Video Group HD - 4x upscaling
Average PSNR· 2016-09-16
36.2
best: 48.23 (RAMS (ours))
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
3DonXiph HD - 4x upscaling
Average PSNR· 2016-09-16
30.3
best: 31.67 (ESPCN)
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
16konSet14 - 4x upscaling
MOS· 2016-09-15
1.8
best: 3.72 (SRGAN)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
16konSet14 - 4x upscaling
SSIM· 2016-09-15
0.7486
best: 0.894 (Edge-informed SR)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
16konBSD100 - 4x upscaling
MOS· 2016-09-15
1.47
best: 3.56 (SRGAN)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
16konBSD100 - 4x upscaling
PSNR· 2016-09-15
25.94
best: 28.16 (DRCT-L)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
16konBSD100 - 4x upscaling
SSIM· 2016-09-15
0.6935
best: 0.851 (Edge-informed SR)
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802

Knowledge Base10 results

2D Human Pose EstimationonVid4 - 4x upscaling
MOVIE· 2016-11-16
9.31
SOTA
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
3D Absolute Human Pose EstimationonVid4 - 4x upscaling
MOVIE· 2016-11-16
9.31
SOTA
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
2D Human Pose EstimationonVid4 - 4x upscaling
PSNR· 2016-11-16
23.82
best: 31.36 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
2D Human Pose EstimationonVid4 - 4x upscaling
SSIM· 2016-11-16
0.6548
best: 0.933 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
3D Absolute Human Pose EstimationonVid4 - 4x upscaling
PSNR· 2016-11-16
23.82
best: 31.36 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
3D Absolute Human Pose EstimationonVid4 - 4x upscaling
SSIM· 2016-11-16
0.6548
best: 0.933 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
2D Human Pose EstimationonUltra Video Group HD - 4x upscaling
Average PSNR· 2016-09-16
36.2
best: 48.23 (RAMS (ours))
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
2D Human Pose EstimationonXiph HD - 4x upscaling
Average PSNR· 2016-09-16
30.3
best: 31.67 (ESPCN)
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
3D Absolute Human Pose EstimationonUltra Video Group HD - 4x upscaling
Average PSNR· 2016-09-16
36.2
best: 48.23 (RAMS (ours))
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
3D Absolute Human Pose EstimationonXiph HD - 4x upscaling
Average PSNR· 2016-09-16
30.3
best: 31.67 (ESPCN)
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158

Playing Games5 results

3D Face AnimationonVid4 - 4x upscaling
MOVIE· 2016-11-16
9.31
SOTA
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
3D Face AnimationonVid4 - 4x upscaling
PSNR· 2016-11-16
23.82
best: 31.36 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
3D Face AnimationonVid4 - 4x upscaling
SSIM· 2016-11-16
0.6548
best: 0.933 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
3D Face AnimationonUltra Video Group HD - 4x upscaling
Average PSNR· 2016-09-16
36.2
best: 48.23 (RAMS (ours))
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
3D Face AnimationonXiph HD - 4x upscaling
Average PSNR· 2016-09-16
30.3
best: 31.67 (ESPCN)
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158

Audio5 results

1 Image, 2*2 StitchionVid4 - 4x upscaling
MOVIE· 2016-11-16
9.31
SOTA
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
1 Image, 2*2 StitchionVid4 - 4x upscaling
PSNR· 2016-11-16
23.82
best: 31.36 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
1 Image, 2*2 StitchionVid4 - 4x upscaling
SSIM· 2016-11-16
0.6548
best: 0.933 (NeuriCam-net)
Real-Time Video Super-Resolution with Spatio-Temporal Networks and Motion Compensation arXiv:1611.05250
1 Image, 2*2 StitchionUltra Video Group HD - 4x upscaling
Average PSNR· 2016-09-16
36.2
best: 48.23 (RAMS (ours))
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158
1 Image, 2*2 StitchionXiph HD - 4x upscaling
Average PSNR· 2016-09-16
30.3
best: 31.67 (ESPCN)
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network arXiv:1609.05158