# Text-and-Video-to-Audio (TV2A) on VGGSound

Source: [Text-and-Video-to-Audio (TV2A) on VGGSound](<https://www.sota2.com/research/sota/text-and-video-to-audio-tv2a-on-vggsound>)

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

Data updated: 2026-06-12T06:47:11.253Z

Evaluation of models performing Text-and-Video-to-Audio tasks (TV2A) on the VGGSound dataset, assessed using metrics such as KL Divergence, Inception Score, Feature Distance, Fréchet Audio Distance, Perceptual Consistency (PC), Perceptual Quality (PQ), and 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 | Text-and-Video-to-Audio (TV2A) |
| Dataset (including any stated split) | VGGSound |
| Metric | KL Divergence |
| 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

- [KL Divergence](<https://www.sota2.com/research/sota/text-and-video-to-audio-tv2a-on-vggsound.md?metric=KL%20Divergence>): Lower is better; unit: SCALAR.

- [Inception Score (IS)](<https://www.sota2.com/research/sota/text-and-video-to-audio-tv2a-on-vggsound.md?metric=Inception%20Score%20(IS)>): Higher is better; unit: SCALAR.

- [Feature Distance (FD)](<https://www.sota2.com/research/sota/text-and-video-to-audio-tv2a-on-vggsound.md?metric=Feature%20Distance%20(FD)>): Lower is better; unit: SCALAR.

- [Fréchet Audio Distance (FAD)](<https://www.sota2.com/research/sota/text-and-video-to-audio-tv2a-on-vggsound.md?metric=Fr%C3%A9chet%20Audio%20Distance%20(FAD)>): Lower is better; unit: SCALAR.

- [Perceptual Consistency (PC)](<https://www.sota2.com/research/sota/text-and-video-to-audio-tv2a-on-vggsound.md?metric=Perceptual%20Consistency%20(PC)>): Higher is better; unit: SCALAR.

- [Perceptual Quality (PQ)](<https://www.sota2.com/research/sota/text-and-video-to-audio-tv2a-on-vggsound.md?metric=Perceptual%20Quality%20(PQ)>): Higher is better; unit: SCALAR.

- [Alignment Score](<https://www.sota2.com/research/sota/text-and-video-to-audio-tv2a-on-vggsound.md?metric=Alignment%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | KL Divergence | Result date | Report paper |
| --- | --- | --- | --- |
| MMAudio | 1.51 | Not recorded | 6a2b926ad33dfb23e783aeac |
| AudioX-Base | 1.56 | Not recorded | 6a2b926ad33dfb23e783aeac |
| AudioX-Turbo | 1.6 | Not recorded | 6a2b926ad33dfb23e783aeac |
| FoleyCrafter | 1.94 | Not recorded | 6a2b926ad33dfb23e783aeac |

### MMAudio

KL Divergence: 1.51. Result ID: 6a2baad3fbc373592c2961cd.

[Reporting paper](<https://www.sota2.com/research/paper/6a2b926ad33dfb23e783aeac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77d71e209131cea75821.md>)

[Implementation](<https://www.sota2.com/hkchengrex.github.io/MMAudio>)

Recorded evaluation configuration: {"Task":"TV2A"}. This record does not assert independent reproduction.

### AudioX-Base

KL Divergence: 1.56. Result ID: 6a2baad3fbc373592c2961ce.

[Reporting paper](<https://www.sota2.com/research/paper/6a2b926ad33dfb23e783aeac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a2b926ad33dfb23e783aeac.md>)

Recorded evaluation configuration: {"Task":"TV2A"}. This record does not assert independent reproduction.

### AudioX-Turbo

KL Divergence: 1.6. Result ID: 6a2baad3fbc373592c2961cf.

[Reporting paper](<https://www.sota2.com/research/paper/6a2b926ad33dfb23e783aeac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a2b926ad33dfb23e783aeac.md>)

Recorded evaluation configuration: {"Task":"TV2A"}. This record does not assert independent reproduction.

### FoleyCrafter

KL Divergence: 1.94. Result ID: 6a2baad3fbc373592c2961cc.

[Reporting paper](<https://www.sota2.com/research/paper/6a2b926ad33dfb23e783aeac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73f01e209131cea5efd0.md>)

Recorded evaluation configuration: {"Task":"TV2A"}. This record does not assert independent reproduction.

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