# Gaussian Grayscale Image Denoising on Urban100

Source: [Gaussian Grayscale Image Denoising on Urban100](<https://www.sota2.com/research/sota/gaussian-grayscale-image-denoising-on-urban100>)

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

Data updated: 2026-02-26T05:57:11.761Z

Evaluation of Gaussian grayscale image denoising performance on the Urban100 dataset, reporting PSNR values specifically for noise levels sigma=15, sigma=25, and sigma=50.

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 | Gaussian grayscale image denoising |
| Dataset (including any stated split) | Urban100 |
| Metric | PSNR (sigma=15) |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| 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

- [PSNR (sigma=15)](<https://www.sota2.com/research/sota/gaussian-grayscale-image-denoising-on-urban100.md?metric=PSNR%20(sigma%3D15)>): Higher is better; unit: SCALAR.

- [PSNR (sigma=25)](<https://www.sota2.com/research/sota/gaussian-grayscale-image-denoising-on-urban100.md?metric=PSNR%20(sigma%3D25)>): Higher is better; unit: SCALAR.

- [PSNR (sigma=50)](<https://www.sota2.com/research/sota/gaussian-grayscale-image-denoising-on-urban100.md?metric=PSNR%20(sigma%3D50)>): Higher is better; unit: SCALAR.

## Recorded results

| Method | PSNR (sigma=15) | Result date | Report paper |
| --- | --- | --- | --- |
| GRL-B+Ours-c | 34.22 | Not recorded | 68fa728e1e209131cea570b8 |
| GRL-B | 34.09 | Not recorded | 68fa728e1e209131cea570b8 |
| Restormer+Ours-c | 33.97 | Not recorded | 68fa728e1e209131cea570b8 |
| Restormer+Ours-c | 33.96 | Not recorded | 68fa728e1e209131cea570b8 |
| Restormer | 33.79 | Not recorded | 68fa728e1e209131cea570b8 |
| Restormer | 33.67 | Not recorded | 68fa728e1e209131cea570b8 |
| DRUNet+Ours-c | 33.65 | Not recorded | 68fa728e1e209131cea570b8 |
| DRUNet | 33.44 | Not recorded | 68fa728e1e209131cea570b8 |

### GRL-B+Ours-c

PSNR (sigma=15): 34.22. Result ID: 69925bce3a6c3148490e6a0a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa728e1e209131cea570b8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa728e1e209131cea570b8.md>)

Recorded evaluation configuration: {"Training protocol":"Separate model per noise level","Pre-trained CLIP model":"true"}. This record does not assert independent reproduction.

### GRL-B

PSNR (sigma=15): 34.09. Result ID: 69925bce3a6c3148490e6a09.

[Reporting paper](<https://www.sota2.com/research/paper/68fa728e1e209131cea570b8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f7b1e209131cea45e06.md>)

[Implementation](<https://github.com/ofsoundof/GRL-Image-Restoration.git>)

Recorded evaluation configuration: {"Training protocol":"Separate model per noise level"}. This record does not assert independent reproduction.

### Restormer+Ours-c

PSNR (sigma=15): 33.97. Result ID: 69925bce3a6c3148490e6a08.

[Reporting paper](<https://www.sota2.com/research/paper/68fa728e1e209131cea570b8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa728e1e209131cea570b8.md>)

Recorded evaluation configuration: {"Training protocol":"Separate model per noise level","Pre-trained CLIP model":"true"}. This record does not assert independent reproduction.

### Restormer+Ours-c

PSNR (sigma=15): 33.96. Result ID: 69925bce3a6c3148490e6a06.

[Reporting paper](<https://www.sota2.com/research/paper/68fa728e1e209131cea570b8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa728e1e209131cea570b8.md>)

Recorded evaluation configuration: {"Training protocol":"Single model for all noise levels","Pre-trained CLIP model":"true"}. This record does not assert independent reproduction.

### Restormer

PSNR (sigma=15): 33.79. Result ID: 69925bce3a6c3148490e6a07.

[Reporting paper](<https://www.sota2.com/research/paper/68fa728e1e209131cea570b8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bfc1e209131cea316e3.md>)

[Implementation](<https://github.com/swz30/Restormer>)

Recorded evaluation configuration: {"Training protocol":"Separate model per noise level"}. This record does not assert independent reproduction.

### Restormer

PSNR (sigma=15): 33.67. Result ID: 69925bce3a6c3148490e6a05.

[Reporting paper](<https://www.sota2.com/research/paper/68fa728e1e209131cea570b8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bfc1e209131cea316e3.md>)

[Implementation](<https://github.com/swz30/Restormer>)

Recorded evaluation configuration: {"Training protocol":"Single model for all noise levels"}. This record does not assert independent reproduction.

### DRUNet+Ours-c

PSNR (sigma=15): 33.65. Result ID: 69925bce3a6c3148490e6a04.

[Reporting paper](<https://www.sota2.com/research/paper/68fa728e1e209131cea570b8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa728e1e209131cea570b8.md>)

Recorded evaluation configuration: {"Training protocol":"Single model for all noise levels","Pre-trained CLIP model":"true"}. This record does not assert independent reproduction.

### DRUNet

PSNR (sigma=15): 33.44. Result ID: 69925bce3a6c3148490e6a03.

[Reporting paper](<https://www.sota2.com/research/paper/68fa728e1e209131cea570b8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a551e209131cea2776a.md>)

Recorded evaluation configuration: {"Training protocol":"Single model for all noise levels"}. This record does not assert independent reproduction.

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