# Text-to-Image Generation on GenEval (InBSim, DrSim, VD/p, CLIP, HPS)

Source: [Text-to-Image Generation on GenEval (InBSim, DrSim, VD/p, CLIP, HPS)](<https://www.sota2.com/research/sota/text-to-image-generation-on-geneval-inbsim-drsim-vd-p-clip-hps>)

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

Data updated: 2026-05-13T08:33:56.568Z

Evaluation of Text-to-Image generation models on the GenEval dataset, focusing on similarity and quality metrics including InBSim, DrSim, VD/p, CLIP score, and HPS.

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 | Text-to-Image Generation |
| Dataset (including any stated split) | GenEval |
| Metric | DrSim |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 15 of 15 |
| 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

- [DrSim](<https://www.sota2.com/research/sota/text-to-image-generation-on-geneval-inbsim-drsim-vd-p-clip-hps.md?metric=DrSim>): Higher is better; unit: SCALAR.

- [InBSim](<https://www.sota2.com/research/sota/text-to-image-generation-on-geneval-inbsim-drsim-vd-p-clip-hps.md?metric=InBSim>): Higher is better; unit: SCALAR.

- [VD/p](<https://www.sota2.com/research/sota/text-to-image-generation-on-geneval-inbsim-drsim-vd-p-clip-hps.md?metric=VD%2Fp>): Higher is better; unit: SCALAR.

- [CLIP Score](<https://www.sota2.com/research/sota/text-to-image-generation-on-geneval-inbsim-drsim-vd-p-clip-hps.md?metric=CLIP%20Score>): Higher is better; unit: SCALAR.

- [HPS](<https://www.sota2.com/research/sota/text-to-image-generation-on-geneval-inbsim-drsim-vd-p-clip-hps.md?metric=HPS>): Higher is better; unit: SCALAR.

## Recorded results

| Method | DrSim | Result date | Report paper |
| --- | --- | --- | --- |
| CADS | 0.634 | Not recorded | 6a0405bcdbc836b87d827702 |
| NoiseOpt\* | 0.446 | Not recorded | 6a0405bcdbc836b87d827702 |
| Input Noise | 0.445 | Not recorded | 6a0405bcdbc836b87d827702 |
| CADS | 0.431 | Not recorded | 6a0405bcdbc836b87d827702 |
| STRIDE | 0.423 | Not recorded | 6a0405bcdbc836b87d827702 |
| GI\* | 0.413 | Not recorded | 6a0405bcdbc836b87d827702 |
| SPELL | 0.369 | Not recorded | 6a0405bcdbc836b87d827702 |
| DDC | 0.368 | Not recorded | 6a0405bcdbc836b87d827702 |
| Baseline | 0.363 | Not recorded | 6a0405bcdbc836b87d827702 |
| STRIDE | 0.337 | Not recorded | 6a0405bcdbc836b87d827702 |
| STRIDE No-PCA | 0.322 | Not recorded | 6a0405bcdbc836b87d827702 |
| Input Noise | 0.314 | Not recorded | 6a0405bcdbc836b87d827702 |
| i.i.d.\* | 0.307 | Not recorded | 6a0405bcdbc836b87d827702 |
| STRIDE No-PCA | 0.274 | Not recorded | 6a0405bcdbc836b87d827702 |
| Baseline | 0.264 | Not recorded | 6a0405bcdbc836b87d827702 |

### CADS

DrSim: 0.634. Result ID: 6a04375047f34ce924ea1e59.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73261e209131cea5ab3e.md>)

[Implementation](<https://github.com/CompVis/latent-diffusion>)

Recorded evaluation configuration: {"Backbone":"SD3.5 Turbo","Denoising Steps":"4-Step","Resolution":"1024x1024"}. This record does not assert independent reproduction.

### NoiseOpt\*

DrSim: 0.446. Result ID: 6a04375047f34ce924ea1e55.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695b457b53fdd5c36174305c.md>)

Recorded evaluation configuration: {"Backbone":"FLUX.1-schnell","Denoising Steps":"1-Step","Resolution":"512x512"}. This record does not assert independent reproduction.

### Input Noise

DrSim: 0.445. Result ID: 6a04375047f34ce924ea1e52.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695b457b53fdd5c36174305c.md>)

Recorded evaluation configuration: {"Backbone":"FLUX.1-schnell","Denoising Steps":"1-Step","Resolution":"512x512"}. This record does not assert independent reproduction.

### CADS

DrSim: 0.431. Result ID: 6a04375047f34ce924ea1e51.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73261e209131cea5ab3e.md>)

[Implementation](<https://github.com/CompVis/latent-diffusion>)

Recorded evaluation configuration: {"Backbone":"FLUX.1-schnell","Denoising Steps":"1-Step","Resolution":"512x512"}. This record does not assert independent reproduction.

### STRIDE

DrSim: 0.423. Result ID: 6a04375047f34ce924ea1e57.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

Recorded evaluation configuration: {"Backbone":"FLUX.1-schnell","Denoising Steps":"1-Step","Resolution":"512x512"}. This record does not assert independent reproduction.

### GI\*

DrSim: 0.413. Result ID: 6a04375047f34ce924ea1e54.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7ad51e209131cea8569a.md>)

Recorded evaluation configuration: {"Backbone":"FLUX.1-schnell","Denoising Steps":"1-Step","Resolution":"512x512"}. This record does not assert independent reproduction.

### SPELL

DrSim: 0.369. Result ID: 6a04375047f34ce924ea1e5c.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78d21e209131cea7a9a2.md>)

Recorded evaluation configuration: {"Backbone":"SD3.5 Turbo","Denoising Steps":"4-Step","Resolution":"1024x1024"}. This record does not assert independent reproduction.

### DDC

DrSim: 0.368. Result ID: 6a04375047f34ce924ea1e5b.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a2d1e209131cea818db.md>)

Recorded evaluation configuration: {"Backbone":"SD3.5 Turbo","Denoising Steps":"4-Step","Resolution":"1024x1024"}. This record does not assert independent reproduction.

### Baseline

DrSim: 0.363. Result ID: 6a04375047f34ce924ea1e50.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

Recorded evaluation configuration: {"Backbone":"FLUX.1-schnell","Denoising Steps":"1-Step","Resolution":"512x512"}. This record does not assert independent reproduction.

### STRIDE

DrSim: 0.337. Result ID: 6a04375047f34ce924ea1e5e.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

Recorded evaluation configuration: {"Backbone":"SD3.5 Turbo","Denoising Steps":"4-Step","Resolution":"1024x1024"}. This record does not assert independent reproduction.

### STRIDE No-PCA

DrSim: 0.322. Result ID: 6a04375047f34ce924ea1e56.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

Recorded evaluation configuration: {"Backbone":"FLUX.1-schnell","Denoising Steps":"1-Step","Resolution":"512x512"}. This record does not assert independent reproduction.

### Input Noise

DrSim: 0.314. Result ID: 6a04375047f34ce924ea1e5a.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695b457b53fdd5c36174305c.md>)

Recorded evaluation configuration: {"Backbone":"SD3.5 Turbo","Denoising Steps":"4-Step","Resolution":"1024x1024"}. This record does not assert independent reproduction.

### i.i.d.\*

DrSim: 0.307. Result ID: 6a04375047f34ce924ea1e53.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695b457b53fdd5c36174305c.md>)

Recorded evaluation configuration: {"Backbone":"FLUX.1-schnell","Denoising Steps":"1-Step","Resolution":"512x512"}. This record does not assert independent reproduction.

### STRIDE No-PCA

DrSim: 0.274. Result ID: 6a04375047f34ce924ea1e5d.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

Recorded evaluation configuration: {"Backbone":"SD3.5 Turbo","Denoising Steps":"4-Step","Resolution":"1024x1024"}. This record does not assert independent reproduction.

### Baseline

DrSim: 0.264. Result ID: 6a04375047f34ce924ea1e58.

[Reporting paper](<https://www.sota2.com/research/paper/6a0405bcdbc836b87d827702.md>)

Recorded evaluation configuration: {"Backbone":"SD3.5 Turbo","Denoising Steps":"4-Step","Resolution":"1024x1024"}. This record does not assert independent reproduction.

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