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/KBNet

KBNet

Reported on 59 benchmarks across 9 tasks · 3 papers · 24 SOTA

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

Computer Vision37 results

Rain RemovalonTest1200
PSNR· 2023-03-06
33.82
best: 34.77 (CAPTNet)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
Rain RemovalonTest1200
SSIM· 2023-03-06
0.931
best: 0.937 (CAPTNet)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
Rain RemovalonTest2800
PSNR· 2023-03-06
34.19
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonSIDD
PSNR (sRGB)· 2023-03-06
40.35
best: 40.39 (CGNet)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonKodak24 sigma50
PSNR· 2023-03-06
30.04
best: 30.13 (DMID-d(MMSE))
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonUrban100 sigma25
PSNR· 2023-03-06
32.96
best: 33.34 (Hi-IR)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonCBSD68 sigma25
PSNR· 2023-03-06
31.8
best: 31.94 (AKDT)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonMcMaster sigma50
PSNR· 2023-03-06
30.27
best: 30.95 (AKDT)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonSet12 sigma50
PSNR· 2023-03-06
28.04
best: 28.15 (Hi-IR)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonSet12 sigma15
PSNR· 2023-03-06
33.4
best: 33.48 (Hi-IR)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonSet12 sigma25
PSNR· 2023-03-06
31.08
best: 31.19 (Hi-IR)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
Image Super-ResolutiononBurstSR
SSIM· 2021-12-14
0.9856
best: 0.987 (BurstM)
SOTA
Kernel-aware Burst Blind Super-Resolution arXiv:2112.07315
3D Object Super-ResolutiononBurstSR
SSIM· 2021-12-14
0.9856
best: 0.987 (BurstM)
SOTA
Kernel-aware Burst Blind Super-Resolution arXiv:2112.07315
Rain RemovalonTest2800
SSIM· 2023-03-06
0.944
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonSIDD
SSIM (sRGB)· 2023-03-06
0.972
best: 0.973 (NBNet)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonCBSD68 sigma15
PSNR· 2023-03-06
34.41
best: 34.64 (AKDT)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
Denoisingonurban100 sigma15
Average PSNR· 2023-03-06
35.15
best: 35.64 (AKDT)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonUrban100 sigma50
PSNR· 2023-03-06
30.04
best: 30.59 (Hi-IR)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonUrban100 sigma25
PSNR· 2023-03-06
31.45
best: 33.34 (Hi-IR)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonUrban100 sigma50
PSNR· 2023-03-06
28.33
best: 30.59 (Hi-IR)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
Denoisingonurban100 sigma15
PSNR· 2023-03-06
33.77
best: 35.63 (AKDT)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonBSD68 sigma15
PSNR· 2023-03-06
31.98
best: 34.01 (RIDNet)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonBSD68 sigma25
PSNR· 2023-03-06
29.54
best: 31.37 (RIDNet)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
DenoisingonBSD68 sigma50
PSNR· 2023-03-06
26.65
best: 30.08 (MaIR+)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
Image Super-ResolutiononBurstSR
LPIPS· 2021-12-14
0.0248
best: 0.021 (BSRT-Large)
Kernel-aware Burst Blind Super-Resolution arXiv:2112.07315
Image Super-ResolutiononBurstSR
PSNR· 2021-12-14
48.27
best: 49.12 (BurstM)
Kernel-aware Burst Blind Super-Resolution arXiv:2112.07315
3D Object Super-ResolutiononBurstSR
LPIPS· 2021-12-14
0.0248
best: 0.021 (BSRT-Large)
Kernel-aware Burst Blind Super-Resolution arXiv:2112.07315
3D Object Super-ResolutiononBurstSR
PSNR· 2021-12-14
48.27
best: 49.12 (BurstM)
Kernel-aware Burst Blind Super-Resolution arXiv:2112.07315
Depth CompletiononKITTI Depth Completion
MAE· 2021-08-24
256.76
best: 199.59 (NLSPN)
Unsupervised Depth Completion with Calibrated Backprojection Layers arXiv:2108.10531
Depth CompletiononKITTI Depth Completion
RMSE· 2021-08-24
1069.47
best: 709.41 (SemAttNet)
Unsupervised Depth Completion with Calibrated Backprojection Layers arXiv:2108.10531
Depth CompletiononKITTI Depth Completion
Runtime [ms]· 2021-08-24
16
best: 220 (NLSPN)
Unsupervised Depth Completion with Calibrated Backprojection Layers arXiv:2108.10531
Depth CompletiononKITTI Depth Completion
iMAE· 2021-08-24
1.02
best: 0.84 (NLSPN)
Unsupervised Depth Completion with Calibrated Backprojection Layers arXiv:2108.10531
Depth CompletiononKITTI Depth Completion
iRMSE· 2021-08-24
2.95
best: 1.99 (NLSPN)
Unsupervised Depth Completion with Calibrated Backprojection Layers arXiv:2108.10531
Depth CompletiononVOID
MAE· 2021-08-24
39.8
best: 26.736 (NLSPN)
Unsupervised Depth Completion with Calibrated Backprojection Layers arXiv:2108.10531
Depth CompletiononVOID
RMSE· 2021-08-24
95.86
best: 79.121 (NLSPN)
Unsupervised Depth Completion with Calibrated Backprojection Layers arXiv:2108.10531
Depth CompletiononVOID
iMAE· 2021-08-24
21.16
best: 12.703 (NLSPN)
Unsupervised Depth Completion with Calibrated Backprojection Layers arXiv:2108.10531
Depth CompletiononVOID
iRMSE· 2021-08-24
49.72
best: 33.876 (NLSPN)
Unsupervised Depth Completion with Calibrated Backprojection Layers arXiv:2108.10531

Adversarial18 results

3D ArchitectureonSIDD
PSNR (sRGB)· 2023-03-06
40.35
best: 40.39 (CGNet)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonKodak24 sigma50
PSNR· 2023-03-06
30.04
best: 30.13 (DMID-d(MMSE))
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonUrban100 sigma25
PSNR· 2023-03-06
32.96
best: 33.34 (Hi-IR)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonCBSD68 sigma25
PSNR· 2023-03-06
31.8
best: 31.94 (AKDT)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonMcMaster sigma50
PSNR· 2023-03-06
30.27
best: 30.95 (AKDT)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonSet12 sigma50
PSNR· 2023-03-06
28.04
best: 28.15 (Hi-IR)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonSet12 sigma15
PSNR· 2023-03-06
33.4
best: 33.48 (Hi-IR)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonSet12 sigma25
PSNR· 2023-03-06
31.08
best: 31.19 (Hi-IR)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonSIDD
SSIM (sRGB)· 2023-03-06
0.972
best: 0.973 (NBNet)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonCBSD68 sigma15
PSNR· 2023-03-06
34.41
best: 34.64 (AKDT)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D Architectureonurban100 sigma15
Average PSNR· 2023-03-06
35.15
best: 35.64 (AKDT)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonUrban100 sigma50
PSNR· 2023-03-06
30.04
best: 30.59 (Hi-IR)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonUrban100 sigma25
PSNR· 2023-03-06
31.45
best: 33.34 (Hi-IR)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonUrban100 sigma50
PSNR· 2023-03-06
28.33
best: 30.59 (Hi-IR)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D Architectureonurban100 sigma15
PSNR· 2023-03-06
33.77
best: 35.63 (AKDT)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonBSD68 sigma15
PSNR· 2023-03-06
31.98
best: 34.01 (RIDNet)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonBSD68 sigma25
PSNR· 2023-03-06
29.54
best: 31.37 (RIDNet)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
3D ArchitectureonBSD68 sigma50
PSNR· 2023-03-06
26.65
best: 30.08 (MaIR+)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881

Graphs3 results

Super-ResolutiononBurstSR
SSIM· 2021-12-14
0.9856
best: 0.987 (BurstM)
SOTA
Kernel-aware Burst Blind Super-Resolution arXiv:2112.07315
Super-ResolutiononBurstSR
LPIPS· 2021-12-14
0.0248
best: 0.021 (BSRT-Large)
Kernel-aware Burst Blind Super-Resolution arXiv:2112.07315
Super-ResolutiononBurstSR
PSNR· 2021-12-14
48.27
best: 49.12 (BurstM)
Kernel-aware Burst Blind Super-Resolution arXiv:2112.07315

Methodology3 results

16konBurstSR
SSIM· 2021-12-14
0.9856
best: 0.987 (BurstM)
SOTA
Kernel-aware Burst Blind Super-Resolution arXiv:2112.07315
16konBurstSR
LPIPS· 2021-12-14
0.0248
best: 0.021 (BSRT-Large)
Kernel-aware Burst Blind Super-Resolution arXiv:2112.07315
16konBurstSR
PSNR· 2021-12-14
48.27
best: 49.12 (BurstM)
Kernel-aware Burst Blind Super-Resolution arXiv:2112.07315

Medical2 results

Image DenoisingonSIDD
PSNR (sRGB)· 2023-03-06
40.35
best: 40.39 (CGNet)
SOTA
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881
Image DenoisingonSIDD
SSIM (sRGB)· 2023-03-06
0.972
best: 0.973 (NBNet)
KBNet: Kernel Basis Network for Image Restoration arXiv:2303.02881