# Video-only Classification on AudioSet

Source: [Video-only Classification on AudioSet](<https://www.sota2.com/research/sota/video-only-classification-on-audioset>)

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

Data updated: 2026-02-26T06:09:32.501Z

Evaluation of model performance using only visual modality inputs for classification tasks on the AudioSet dataset.

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 | Video-only Classification |
| Dataset (including any stated split) | AudioSet |
| Metric | mAP |
| Unit | PERCENT |
| 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

- [mAP](<https://www.sota2.com/research/sota/video-only-classification-on-audioset.md?metric=mAP>): Higher is better; unit: PERCENT.

## Recorded results

| Method | mAP | Result date | Report paper |
| --- | --- | --- | --- |
| MBT | 31.3 | Not recorded | 68fa71ba1e209131cea52cb5 |
| AV-MAE | 31.1 | Not recorded | 68fa71ba1e209131cea52cb5 |
| Perceiver | 25.8 | Not recorded | 68fa71ba1e209131cea52cb5 |
| Fayek et al. | 25.7 | Not recorded | 68fa71ba1e209131cea52cb5 |
| GBlend | 18.8 | Not recorded | 68fa71ba1e209131cea52cb5 |

### MBT

mAP: 31.3. Result ID: 6991dd713a6c3148490b1587.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cb5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e221e209131cea3e0de.md>)

Recorded evaluation configuration: {"Pretraining":"Im21K","Training set":"AS-500K"}. This record does not assert independent reproduction.

### AV-MAE

mAP: 31.1. Result ID: 6991dd713a6c3148490b1588.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cb5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cb5.md>)

[Implementation](<https://github.com/google-research/scenic>)

Recorded evaluation configuration: {"Pretraining":"SSL AS-2M","Training set":"AS-500K"}. This record does not assert independent reproduction.

### Perceiver

mAP: 25.8. Result ID: 6991dd713a6c3148490b1585.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cb5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a381e209131cea26a64.md>)

Recorded evaluation configuration: {"Pretraining":"None","Training set":"AS-2M"}. This record does not assert independent reproduction.

### Fayek et al.

mAP: 25.7. Result ID: 6991dd713a6c3148490b1586.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cb5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67b11e209131cea17da6.md>)

Recorded evaluation configuration: {"Pretraining":"Im1K","Training set":"AS-2M"}. This record does not assert independent reproduction.

### GBlend

mAP: 18.8. Result ID: 6991dd713a6c3148490b1584.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cb5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67561e209131cea15a42.md>)

Recorded evaluation configuration: {"Pretraining":"Im1K","Training set":"AS-2M"}. This record does not assert independent reproduction.

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