# Image Reconstruction on ImageNet 1024x1024 (val)

Source: [Image Reconstruction on ImageNet 1024x1024 (val)](<https://www.sota2.com/research/sota/image-reconstruction-on-imagenet-1024x1024-val>)

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

Data updated: 2026-06-01T08:34:35.786Z

Evaluation of image reconstruction models using the validation split of ImageNet rendered at 1024x1024 resolution, measured using rFID, PSNR, and SSIM.

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 Reconstruction |
| Dataset (including any stated split) | ImageNet 1024x1024 (val) |
| Metric | rFID |
| 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

- [rFID](<https://www.sota2.com/research/sota/image-reconstruction-on-imagenet-1024x1024-val.md?metric=rFID>): Lower is better; unit: SCALAR.

- [PSNR](<https://www.sota2.com/research/sota/image-reconstruction-on-imagenet-1024x1024-val.md?metric=PSNR>): Higher is better; unit: SCALAR.

- [SSIM](<https://www.sota2.com/research/sota/image-reconstruction-on-imagenet-1024x1024-val.md?metric=SSIM>): Higher is better; unit: PERCENT.

## Recorded results

| Method | rFID | Result date | Report paper |
| --- | --- | --- | --- |
| MergeTok-BL (VQ) | 1.87 | Not recorded | 6a1d11e1dbd7150506f83dd6 |
| TiTok-S-128 | 2.32 | Not recorded | 6a1d11e1dbd7150506f83dd6 |
| Layton-H | 2.78 | Not recorded | 6a1d11e1dbd7150506f83dd6 |
| LlamaGen-Tok. | 3.17 | Not recorded | 6a1d11e1dbd7150506f83dd6 |
| OpenMAGVIT2 | 3.43 | Not recorded | 6a1d11e1dbd7150506f83dd6 |
| Taming-VQGAN | 6.02 | Not recorded | 6a1d11e1dbd7150506f83dd6 |

### MergeTok-BL (VQ)

rFID: 1.87. Result ID: 6a1d4329fbc373592c16747a.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d11e1dbd7150506f83dd6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1d11e1dbd7150506f83dd6.md>)

Recorded evaluation configuration: {"Type":"1D VAE+VQ","Ratio":"16","#Tok.":"256","#Code":"2^14"}. This record does not assert independent reproduction.

### TiTok-S-128

rFID: 2.32. Result ID: 6a1d4329fbc373592c167478.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d11e1dbd7150506f83dd6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73991e209131cea5d7e1.md>)

[Implementation](<https://yucornetto.github.io/projects/titok.html>)

Recorded evaluation configuration: {"Type":"1D VQ","Ratio":"32","#Tok.":"128","#Code":"2^12"}. This record does not assert independent reproduction.

### Layton-H

rFID: 2.78. Result ID: 6a1d4329fbc373592c167479.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d11e1dbd7150506f83dd6.md>)

Recorded evaluation configuration: {"Type":"1D Diff.","Ratio":"32","#Tok.":"-","#Code":"-"}. This record does not assert independent reproduction.

### LlamaGen-Tok.

rFID: 3.17. Result ID: 6a1d4329fbc373592c167477.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d11e1dbd7150506f83dd6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa738a1e209131cea5d111.md>)

[Implementation](<https://github.com/FoundationVision/LlamaGen>)

Recorded evaluation configuration: {"Type":"2D VQ","Ratio":"32","#Tok.":"8^2","#Code":"2^14"}. This record does not assert independent reproduction.

### OpenMAGVIT2

rFID: 3.43. Result ID: 6a1d4329fbc373592c167476.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d11e1dbd7150506f83dd6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76e01e209131cea6f87d.md>)

[Implementation](<https://github.com/TencentARC/Open-MAGVIT2>)

Recorded evaluation configuration: {"Type":"2D LFQ","Ratio":"16","#Tok.":"16^2","#Code":"2^18"}. This record does not assert independent reproduction.

### Taming-VQGAN

rFID: 6.02. Result ID: 6a1d4329fbc373592c167475.

[Reporting paper](<https://www.sota2.com/research/paper/6a1d11e1dbd7150506f83dd6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a371e209131cea26a4d.md>)

[Implementation](<https://git.io/JnyvK>)

Recorded evaluation configuration: {"Type":"2D VQ","Ratio":"16","#Tok.":"16^2","#Code":"2^10"}. This record does not assert independent reproduction.

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