# Image Extension on Places365 held-out set of 500 images (50% extension)

Source: [Image Extension on Places365 held-out set of 500 images (50% extension)](<https://www.sota2.com/research/sota/image-extension-on-places365-held-out-set-of-500-images-50-percent-extension>)

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

Data updated: 2026-02-26T05:44:10.765Z

Evaluation of image extension quality on a specific held-out subset of the Places365 dataset, typically measured using FID and PSNR.

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 Extension |
| Dataset (including any stated split) | Places365 held-out set of 500 images (50% extension) |
| Metric | FID |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 4 of 4 |
| 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/image-extension-on-places365-held-out-set-of-500-images-50-percent-extension.md?metric=FID>): Lower is better; unit: SCALAR.

- [PSNR](<https://www.sota2.com/research/sota/image-extension-on-places365-held-out-set-of-500-images-50-percent-extension.md?metric=PSNR>): Higher is better; unit: SCALAR.

## Recorded results

| Method | FID | Result date | Report paper |
| --- | --- | --- | --- |
| Boundless | 3.46 | Not recorded | 68fa66371e209131cea0ed63 |
| Boundless (No-Cond) | 5.01 | Not recorded | 68fa66371e209131cea0ed63 |
| PartialConv | 11.2 | Not recorded | 68fa66371e209131cea0ed63 |
| DeepFillv2 | 11.65 | Not recorded | 68fa66371e209131cea0ed63 |

### Boundless

FID: 3.46. Result ID: 698ce0ec8b517eb4331d62b3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66371e209131cea0ed63.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66371e209131cea0ed63.md>)

Recorded evaluation configuration: {"Mask ratio":"50% extension"}. This record does not assert independent reproduction.

### Boundless (No-Cond)

FID: 5.01. Result ID: 698ce0ec8b517eb4331d62b2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66371e209131cea0ed63.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66371e209131cea0ed63.md>)

Recorded evaluation configuration: {"Mask ratio":"50% extension","Discriminator conditioning":false}. This record does not assert independent reproduction.

### PartialConv

FID: 11.2. Result ID: 698ce0ec8b517eb4331d62b1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66371e209131cea0ed63.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa652e1e209131cea08a58.md>)

Recorded evaluation configuration: {"Mask ratio":"50% extension"}. This record does not assert independent reproduction.

### DeepFillv2

FID: 11.65. Result ID: 698ce0ec8b517eb4331d62b0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66371e209131cea0ed63.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66941e209131cea1110d.md>)

[Implementation](<https://github.com/JiahuiYu/generative_inpainting>)

Recorded evaluation configuration: {"Mask ratio":"50% extension"}. This record does not assert independent reproduction.

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