# Veiling Glare Removal on Realworld-MRL (test)

Source: [Veiling Glare Removal on Realworld-MRL (test)](<https://www.sota2.com/research/sota/veiling-glare-removal-on-realworld-mrl-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-09T06:59:19.850Z

Evaluation of Veiling Glare Removal performance on the test split of the Realworld-MRL dataset, measured using CLIPIQA, Q-Align, and NIQE.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Veiling glare removal |
| Dataset (including any stated split) | Realworld-MRL (test) |
| Metric | CLIPIQA |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 12 of 12 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [CLIPIQA](<https://www.sota2.com/research/sota/veiling-glare-removal-on-realworld-mrl-test.md?metric=CLIPIQA>): Higher is better; unit: SCALAR.

- [Q-Align](<https://www.sota2.com/research/sota/veiling-glare-removal-on-realworld-mrl-test.md?metric=Q-Align>): Higher is better; unit: SCALAR.

- [NIQE](<https://www.sota2.com/research/sota/veiling-glare-removal-on-realworld-mrl-test.md?metric=NIQE>): Lower is better; unit: SCALAR.

## Recorded results

| Method | CLIPIQA | Result date | Report paper |
| --- | --- | --- | --- |
| DeVeiler | 0.44 | Not recorded | 6923e675d6b0dd334292eed7 |
| SwinIR+DiffDehaze | 0.437 | Not recorded | 6923e675d6b0dd334292eed7 |
| DiffDehaze | 0.428 | Not recorded | 6923e675d6b0dd334292eed7 |
| CycleGAN | 0.418 | Not recorded | 6923e675d6b0dd334292eed7 |
| SwinIR+RIDCP | 0.399 | Not recorded | 6923e675d6b0dd334292eed7 |
| NAFNET | 0.398 | Not recorded | 6923e675d6b0dd334292eed7 |
| Fast two-step | 0.388 | Not recorded | 6923e675d6b0dd334292eed7 |
| SwinIR+Flare7K++ | 0.387 | Not recorded | 6923e675d6b0dd334292eed7 |
| QDMR | 0.376 | Not recorded | 6923e675d6b0dd334292eed7 |
| SwinIR | 0.374 | Not recorded | 6923e675d6b0dd334292eed7 |
| DiffBIR | 0.367 | Not recorded | 6923e675d6b0dd334292eed7 |
| UCL-Dehaze | 0.343 | Not recorded | 6923e675d6b0dd334292eed7 |

### DeVeiler

CLIPIQA: 0.44. Result ID: 69ae6f286dbb7075b955fb82.

[Reporting paper](<https://www.sota2.com/research/paper/6923e675d6b0dd334292eed7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6923e675d6b0dd334292eed7.md>)

Recorded evaluation configuration: {"Pipeline Category":"Domain Adaptation Pipeline","Pre-trained status":"Proposed method"}. This record does not assert independent reproduction.

### SwinIR+DiffDehaze

CLIPIQA: 0.437. Result ID: 69ae6f286dbb7075b955fb7c.

[Reporting paper](<https://www.sota2.com/research/paper/6923e675d6b0dd334292eed7.md>)

Recorded evaluation configuration: {"Pipeline Category":"Cascaded Pipeline","Pre-trained status":"Pre-trained model used"}. This record does not assert independent reproduction.

### DiffDehaze

CLIPIQA: 0.428. Result ID: 69ae6f286dbb7075b955fb80.

[Reporting paper](<https://www.sota2.com/research/paper/6923e675d6b0dd334292eed7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77a41e209131cea74912.md>)

Recorded evaluation configuration: {"Pipeline Category":"Domain Adaptation Pipeline","Pre-trained status":"Retrained"}. This record does not assert independent reproduction.

### CycleGAN

CLIPIQA: 0.418. Result ID: 69ae6f286dbb7075b955fb7e.

[Reporting paper](<https://www.sota2.com/research/paper/6923e675d6b0dd334292eed7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68351e209131cea1ac7a.md>)

Recorded evaluation configuration: {"Pipeline Category":"Domain Adaptation Pipeline","Pre-trained status":"Retrained"}. This record does not assert independent reproduction.

### SwinIR+RIDCP

CLIPIQA: 0.399. Result ID: 69ae6f286dbb7075b955fb7b.

[Reporting paper](<https://www.sota2.com/research/paper/6923e675d6b0dd334292eed7.md>)

Recorded evaluation configuration: {"Pipeline Category":"Cascaded Pipeline","Pre-trained status":"Pre-trained model used"}. This record does not assert independent reproduction.

### NAFNET

CLIPIQA: 0.398. Result ID: 69ae6f286dbb7075b955fb79.

[Reporting paper](<https://www.sota2.com/research/paper/6923e675d6b0dd334292eed7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d1d1e209131cea37def.md>)

[Implementation](<https://www.sota2.com/github.com/megvii-research/NAFNet>)

Recorded evaluation configuration: {"Pipeline Category":"Single Degradation Pipeline","Pre-trained status":"Retrained"}. This record does not assert independent reproduction.

### Fast two-step

CLIPIQA: 0.388. Result ID: 69ae6f286dbb7075b955fb77.

[Reporting paper](<https://www.sota2.com/research/paper/6923e675d6b0dd334292eed7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d1d1e209131cea37db3.md>)

Recorded evaluation configuration: {"Pipeline Category":"Single Degradation Pipeline","Pre-trained status":"Retrained"}. This record does not assert independent reproduction.

### SwinIR+Flare7K++

CLIPIQA: 0.387. Result ID: 69ae6f286dbb7075b955fb7d.

[Reporting paper](<https://www.sota2.com/research/paper/6923e675d6b0dd334292eed7.md>)

Recorded evaluation configuration: {"Pipeline Category":"Cascaded Pipeline","Pre-trained status":"Pre-trained model used"}. This record does not assert independent reproduction.

### QDMR

CLIPIQA: 0.376. Result ID: 69ae6f286dbb7075b955fb81.

[Reporting paper](<https://www.sota2.com/research/paper/6923e675d6b0dd334292eed7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa487c1e209131ce92ff57.md>)

Recorded evaluation configuration: {"Pipeline Category":"Domain Adaptation Pipeline","Pre-trained status":"Retrained"}. This record does not assert independent reproduction.

### SwinIR

CLIPIQA: 0.374. Result ID: 69ae6f286dbb7075b955fb78.

[Reporting paper](<https://www.sota2.com/research/paper/6923e675d6b0dd334292eed7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a9a1e209131cea294aa.md>)

[Implementation](<https://github.com/JingyunLiang/SwinIR>)

Recorded evaluation configuration: {"Pipeline Category":"Single Degradation Pipeline","Pre-trained status":"Retrained"}. This record does not assert independent reproduction.

### DiffBIR

CLIPIQA: 0.367. Result ID: 69ae6f286dbb7075b955fb7a.

[Reporting paper](<https://www.sota2.com/research/paper/6923e675d6b0dd334292eed7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72e01e209131cea5918f.md>)

[Implementation](<https://github.com/XPixelGroup/DiffBIR>)

Recorded evaluation configuration: {"Pipeline Category":"Single Degradation Pipeline","Pre-trained status":"Retrained"}. This record does not assert independent reproduction.

### UCL-Dehaze

CLIPIQA: 0.343. Result ID: 69ae6f286dbb7075b955fb7f.

[Reporting paper](<https://www.sota2.com/research/paper/6923e675d6b0dd334292eed7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c7f1e209131cea345a8.md>)

Recorded evaluation configuration: {"Pipeline Category":"Domain Adaptation Pipeline","Pre-trained status":"Retrained"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
