# Unconditional Image Generation on CelebA 128x128 (train test)

Source: [Unconditional Image Generation on CelebA 128x128 (train test)](<https://www.sota2.com/research/sota/unconditional-image-generation-on-celeba-128x128-train-test>)

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

Data updated: 2026-02-26T05:44:12.389Z

Evaluation of unconditional image generation models on the combined train and test splits of the CelebA dataset rendered at 128x128 resolution, typically measured using FID and IS.

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 | Unconditional Image Generation |
| Dataset (including any stated split) | CelebA 128x128 (train test) |
| Metric | FID |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 6 of 6 |
| 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

- [FID](<https://www.sota2.com/research/sota/unconditional-image-generation-on-celeba-128x128-train-test.md?metric=FID>): Lower is better; unit: SCALAR.

- [IS](<https://www.sota2.com/research/sota/unconditional-image-generation-on-celeba-128x128-train-test.md?metric=IS>): Higher is better; unit: SCALAR.

## Recorded results

| Method | FID | Result date | Report paper |
| --- | --- | --- | --- |
| U-Net GAN | 2.03 | Not recorded | 68fa69971e209131cea22eac |
| U-Net GAN | 2.95 | Not recorded | 68fa69971e209131cea22eac |
| BigGAN | 3.7 | Not recorded | 68fa69971e209131cea22eac |
| BigGAN | 4.54 | Not recorded | 68fa69971e209131cea22eac |
| COCO-GAN | 5.74 | Not recorded | 68fa69971e209131cea22eac |
| PG-GAN | 7.3 | Not recorded | 68fa69971e209131cea22eac |

### U-Net GAN

FID: 2.03. Result ID: 698dbd8f8b517eb433237162.

[Reporting paper](<https://www.sota2.com/research/paper/68fa69971e209131cea22eac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69971e209131cea22eac.md>)

[Implementation](<https://github.com/boschresearch/unetgan>)

Recorded evaluation configuration: {"Framework":"PyTorch","Resolution":"128x128","Training Iterations":"800k"}. This record does not assert independent reproduction.

### U-Net GAN

FID: 2.95. Result ID: 698dbd8f8b517eb433237164.

[Reporting paper](<https://www.sota2.com/research/paper/68fa69971e209131cea22eac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69971e209131cea22eac.md>)

[Implementation](<https://github.com/boschresearch/unetgan>)

Recorded evaluation configuration: {"Framework":"TensorFlow","Resolution":"128x128","Training Iterations":"800k"}. This record does not assert independent reproduction.

### BigGAN

FID: 3.7. Result ID: 698dbd8f8b517eb433237161.

[Reporting paper](<https://www.sota2.com/research/paper/68fa69971e209131cea22eac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65661e209131cea0a090.md>)

[Implementation](<https://tfhub.dev/s?q=biggan>)

Recorded evaluation configuration: {"Framework":"PyTorch","Resolution":"128x128","Training Iterations":"800k"}. This record does not assert independent reproduction.

### BigGAN

FID: 4.54. Result ID: 698dbd8f8b517eb433237163.

[Reporting paper](<https://www.sota2.com/research/paper/68fa69971e209131cea22eac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65661e209131cea0a090.md>)

[Implementation](<https://tfhub.dev/s?q=biggan>)

Recorded evaluation configuration: {"Framework":"TensorFlow","Resolution":"128x128","Training Iterations":"800k"}. This record does not assert independent reproduction.

### COCO-GAN

FID: 5.74. Result ID: 698dbd8f8b517eb433237160.

[Reporting paper](<https://www.sota2.com/research/paper/68fa69971e209131cea22eac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66e81e209131cea13113.md>)

[Implementation](<https://github.com/bioinf-jku/TTUR>)

Recorded evaluation configuration: {"Framework":"TensorFlow","Resolution":"128x128"}. This record does not assert independent reproduction.

### PG-GAN

FID: 7.3. Result ID: 698dbd8f8b517eb43323715f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa69971e209131cea22eac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64151e209131cea02a8e.md>)

[Implementation](<https://github.com/tkarras/progressive_growing_of_gans>)

Recorded evaluation configuration: {"Framework":"TensorFlow","Resolution":"128x128"}. This record does not assert independent reproduction.

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