# Face Stylization on Prev Style Images (low-density)

Source: [Face Stylization on Prev Style Images (low-density)](<https://www.sota2.com/research/sota/face-stylization-on-prev-style-images-low-density>)

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

Data updated: 2026-02-26T05:35:02.165Z

Evaluation of Face Stylization models using the low-density subset of the Prev Style Images dataset, measured using ArtFID, ID Similarity, and Structure Distance.

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 | Face Stylization |
| Dataset (including any stated split) | Prev Style Images (low-density) |
| Metric | ArtFID |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 5 of 5 |
| 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

- [ArtFID](<https://www.sota2.com/research/sota/face-stylization-on-prev-style-images-low-density.md?metric=ArtFID>): Lower is better; unit: SCALAR.

- [ID Similarity](<https://www.sota2.com/research/sota/face-stylization-on-prev-style-images-low-density.md?metric=ID%20Similarity>): Higher is better; unit: SCALAR.

- [Structure Distance](<https://www.sota2.com/research/sota/face-stylization-on-prev-style-images-low-density.md?metric=Structure%20Distance>): Lower is better; unit: SCALAR.

## Recorded results

| Method | ArtFID | Result date | Report paper |
| --- | --- | --- | --- |
| OSASIS | 33.2 | Not recorded | 68fa72c11e209131cea58692 |
| InST | 35.86 | Not recorded | 68fa72c11e209131cea58692 |
| MTG+HFGI | 37.27 | Not recorded | 68fa72c11e209131cea58692 |
| JoJoGAN+HFGI | 41.09 | Not recorded | 68fa72c11e209131cea58692 |
| DiffuseIT | 48.18 | Not recorded | 68fa72c11e209131cea58692 |

### OSASIS

ArtFID: 33.2. Result ID: 699279033a6c3148490f3269.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72c11e209131cea58692.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72c11e209131cea58692.md>)

[Implementation](<https://github.com/hansam95/OSASIS>)

Recorded evaluation configuration: {"Pre-training dataset":"FFHQ"}. This record does not assert independent reproduction.

### InST

ArtFID: 35.86. Result ID: 699279033a6c3148490f3268.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72c11e209131cea58692.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ecf1e209131cea4250e.md>)

[Implementation](<https://github.com/zyxElsa/InST>)

Recorded evaluation configuration: {"Pre-training dataset":"FFHQ"}. This record does not assert independent reproduction.

### MTG+HFGI

ArtFID: 37.27. Result ID: 699279033a6c3148490f3265.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72c11e209131cea58692.md>)

Recorded evaluation configuration: {"Inversion method":"HFGI","Pre-training dataset":"FFHQ"}. This record does not assert independent reproduction.

### JoJoGAN+HFGI

ArtFID: 41.09. Result ID: 699279033a6c3148490f3266.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72c11e209131cea58692.md>)

Recorded evaluation configuration: {"Inversion method":"HFGI","Pre-training dataset":"FFHQ"}. This record does not assert independent reproduction.

### DiffuseIT

ArtFID: 48.18. Result ID: 699279033a6c3148490f3267.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72c11e209131cea58692.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e731e209131cea3fe10.md>)

Recorded evaluation configuration: {"Pre-training dataset":"FFHQ"}. This record does not assert independent reproduction.

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