# Unconditional Co-generation on Image-Text-Audio

Source: [Unconditional Co-generation on Image-Text-Audio](<https://www.sota2.com/research/sota/unconditional-co-generation-on-image-text-audio>)

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

Data updated: 2026-06-16T10:02:05.153Z

Evaluation of models performing unconditional co-generation across modalities (Text, Image, Audio) using the Image-Text-Audio dataset, measured using CLIP, CLAP, and AIS alignment scores.

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 co-generation |
| Dataset (including any stated split) | image-text-audio |
| Metric | CLIP Score (Text-Image) |
| Unit | SCALAR |
| Direction | Higher 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

- [CLIP Score (Text-Image)](<https://www.sota2.com/research/sota/unconditional-co-generation-on-image-text-audio.md?metric=CLIP%20Score%20(Text-Image)>): Higher is better; unit: SCALAR.

- [CLAP Score (Text-Audio)](<https://www.sota2.com/research/sota/unconditional-co-generation-on-image-text-audio.md?metric=CLAP%20Score%20(Text-Audio)>): Higher is better; unit: SCALAR.

- [AIS Score (Audio-Image)](<https://www.sota2.com/research/sota/unconditional-co-generation-on-image-text-audio.md?metric=AIS%20Score%20(Audio-Image)>): Higher is better; unit: SCALAR.

## Recorded results

| Method | CLIP Score (Text-Image) | Result date | Report paper |
| --- | --- | --- | --- |
| MUNI | 26.76 | Not recorded | 6a30d90cd33dfb23e783d228 |
| MUNI† | 26.74 | Not recorded | 6a30d90cd33dfb23e783d228 |
| MoPoE | 25.83 | Not recorded | 6a30d90cd33dfb23e783d228 |
| MMVAE | 25.52 | Not recorded | 6a30d90cd33dfb23e783d228 |
| OmniFlow | 21.17 | Not recorded | 6a30d90cd33dfb23e783d228 |

### MUNI

CLIP Score (Text-Image): 26.76. Result ID: 6a311f08fbc373592c2e087d.

[Reporting paper](<https://www.sota2.com/research/paper/6a30d90cd33dfb23e783d228.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a30d90cd33dfb23e783d228.md>)

Recorded evaluation configuration: {"Category":"Generalists","Aggregation":"PoE"}. This record does not assert independent reproduction.

### MUNI†

CLIP Score (Text-Image): 26.74. Result ID: 6a311f08fbc373592c2e087e.

[Reporting paper](<https://www.sota2.com/research/paper/6a30d90cd33dfb23e783d228.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a30d90cd33dfb23e783d228.md>)

Recorded evaluation configuration: {"Category":"Generalists","Aggregation":"Hellinger"}. This record does not assert independent reproduction.

### MoPoE

CLIP Score (Text-Image): 25.83. Result ID: 6a311f08fbc373592c2e0880.

[Reporting paper](<https://www.sota2.com/research/paper/6a30d90cd33dfb23e783d228.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a391e209131cea26c1e.md>)

Recorded evaluation configuration: {"Category":"Multimodal VAEs"}. This record does not assert independent reproduction.

### MMVAE

CLIP Score (Text-Image): 25.52. Result ID: 6a311f08fbc373592c2e087f.

[Reporting paper](<https://www.sota2.com/research/paper/6a30d90cd33dfb23e783d228.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66991e209131cea113ec.md>)

Recorded evaluation configuration: {"Category":"Multimodal VAEs"}. This record does not assert independent reproduction.

### OmniFlow

CLIP Score (Text-Image): 21.17. Result ID: 6a311f08fbc373592c2e087c.

[Reporting paper](<https://www.sota2.com/research/paper/6a30d90cd33dfb23e783d228.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77911e209131cea73fd3.md>)

[Implementation](<https://github.com/jacklishufan/OmniFlows>)

Recorded evaluation configuration: {"Category":"Generalists"}. This record does not assert independent reproduction.

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