SOTAVerified

Image Super-Resolution

Image Super-Resolution is a machine learning task where the goal is to increase the resolution of an image, often by a factor of 4x or more, while maintaining its content and details as much as possible. The end result is a high-resolution version of the original image. This task can be used for various applications such as improving image quality, enhancing visual detail, and increasing the accuracy of computer vision algorithms.

Papers

Showing 1–10 of 1589 papers

TitleStatusHype
SpectraLift: Physics-Guided Spectral-Inversion Network for Self-Supervised Hyperspectral Image Super-Resolution—0
IM-LUT: Interpolation Mixing Look-Up Tables for Image Super-ResolutionCode1
Efficient Feedback Gate Network for Hyperspectral Image Super-Resolution—0
Unsupervised Image Super-Resolution Reconstruction Based on Real-World Degradation Patterns—0
Efficient Star Distillation Attention Network for Lightweight Image Super-Resolution—0
Structural Similarity-Inspired Unfolding for Lightweight Image Super-ResolutionCode1
Stroke-based Cyclic Amplifier: Image Super-Resolution at Arbitrary Ultra-Large Scales—0
Incorporating Uncertainty-Guided and Top-k Codebook Matching for Real-World Blind Image Super-Resolution—0
Task-driven real-world super-resolution of document scans—0
Practical Manipulation Model for Robust Deepfake DetectionCode0
Show:102550
← PrevPage 1 of 159Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1DRCT-LPSNR35.36—Unverified
2HMA†PSNR35.33—Unverified
3HAT-LPSNR35.29—Unverified
4Hi-IR-LPSNR35.27—Unverified
5HAT_FIRPSNR35.17—Unverified
6HATPSNR35.13—Unverified
7CPAT+PSNR34.97—Unverified
8DRCTPSNR34.96—Unverified
9SwinFIRPSNR34.93—Unverified
10CPATPSNR34.91—Unverified