# Image Style Transfer on OmniStyle 150K (held-out)

Source: [Image Style Transfer on OmniStyle 150K (held-out)](<https://www.sota2.com/research/sota/image-style-transfer-on-omnistyle-150k-held-out>)

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

Data updated: 2026-04-24T07:10:58.321Z

Evaluation of Image Style Transfer performance on the OmniStyle 150K held-out set, measured using perceptual metrics (LPIPS, SSIM, CLIP Sim, DreamSim) and component losses (Style Loss, Content Loss).

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 | Image Style Transfer |
| Dataset (including any stated split) | OmniStyle 150K (held-out) |
| Metric | Style Loss |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 3 of 3 |
| 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

- [Style Loss](<https://www.sota2.com/research/sota/image-style-transfer-on-omnistyle-150k-held-out.md?metric=Style%20Loss>): Lower is better; unit: SCALAR.

- [Content Loss](<https://www.sota2.com/research/sota/image-style-transfer-on-omnistyle-150k-held-out.md?metric=Content%20Loss>): Lower is better; unit: SCALAR.

- [LPIPS](<https://www.sota2.com/research/sota/image-style-transfer-on-omnistyle-150k-held-out.md?metric=LPIPS>): Higher is better; unit: SCALAR.

- [SSIM](<https://www.sota2.com/research/sota/image-style-transfer-on-omnistyle-150k-held-out.md?metric=SSIM>): Higher is better; unit: SCALAR.

- [DreamSim](<https://www.sota2.com/research/sota/image-style-transfer-on-omnistyle-150k-held-out.md?metric=DreamSim>): Higher is better; unit: SCALAR.

- [CLIP Sim](<https://www.sota2.com/research/sota/image-style-transfer-on-omnistyle-150k-held-out.md?metric=CLIP%20Sim>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Style Loss | Result date | Report paper |
| --- | --- | --- | --- |
| StyleVAR | 0.0466 | Not recorded | 69eaf8e355a3c32af89f8788 |
| StyleVAR | 0.0468 | Not recorded | 69eaf8e355a3c32af89f8788 |
| AdaIN | 0.0625 | Not recorded | 69eaf8e355a3c32af89f8788 |

### StyleVAR

Style Loss: 0.0466. Result ID: 69eb177647f34ce924d2cf95.

[Reporting paper](<https://www.sota2.com/research/paper/69eaf8e355a3c32af89f8788.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69eaf8e355a3c32af89f8788.md>)

Recorded evaluation configuration: {"variant":"GRPO","Infer (s)":"0.4031"}. This record does not assert independent reproduction.

### StyleVAR

Style Loss: 0.0468. Result ID: 69eb177647f34ce924d2cf94.

[Reporting paper](<https://www.sota2.com/research/paper/69eaf8e355a3c32af89f8788.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69eaf8e355a3c32af89f8788.md>)

Recorded evaluation configuration: {"variant":"SFT","Infer (s)":"0.4031"}. This record does not assert independent reproduction.

### AdaIN

Style Loss: 0.0625. Result ID: 69eb177647f34ce924d2cf93.

[Reporting paper](<https://www.sota2.com/research/paper/69eaf8e355a3c32af89f8788.md>)

Recorded evaluation configuration: {"variant":"baseline","Infer (s)":"0.0079"}. This record does not assert independent reproduction.

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