# Audio-visual video parsing on LLP (val)

Source: [Audio-visual video parsing on LLP (val)](<https://www.sota2.com/research/sota/audio-visual-video-parsing-on-llp-val>)

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

Data updated: 2026-05-12T09:30:58.015Z

Evaluation of Audio-visual video parsing performance on the validation split of the LLP dataset, reporting metrics across both segment-level and event-level analysis for Audio, Visual, Audio-Visual, Type, and Event components.

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 | Audio-visual video parsing |
| Dataset (including any stated split) | LLP (val) |
| Metric | Segment-Level Audio Accuracy |
| Unit | PERCENT |
| Direction | Higher 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

- [Segment-Level Audio Accuracy](<https://www.sota2.com/research/sota/audio-visual-video-parsing-on-llp-val.md?metric=Segment-Level%20Audio%20Accuracy>): Higher is better; unit: PERCENT.

- [Segment-Level Visual Accuracy](<https://www.sota2.com/research/sota/audio-visual-video-parsing-on-llp-val.md?metric=Segment-Level%20Visual%20Accuracy>): Higher is better; unit: PERCENT.

- [Segment-Level Audio-Visual Accuracy](<https://www.sota2.com/research/sota/audio-visual-video-parsing-on-llp-val.md?metric=Segment-Level%20Audio-Visual%20Accuracy>): Higher is better; unit: PERCENT.

- [Segment-Level Type Accuracy](<https://www.sota2.com/research/sota/audio-visual-video-parsing-on-llp-val.md?metric=Segment-Level%20Type%20Accuracy>): Higher is better; unit: PERCENT.

- [Segment-Level Event Accuracy](<https://www.sota2.com/research/sota/audio-visual-video-parsing-on-llp-val.md?metric=Segment-Level%20Event%20Accuracy>): Higher is better; unit: PERCENT.

- [Event-Level Audio Accuracy](<https://www.sota2.com/research/sota/audio-visual-video-parsing-on-llp-val.md?metric=Event-Level%20Audio%20Accuracy>): Higher is better; unit: PERCENT.

- [Event-Level Visual Accuracy](<https://www.sota2.com/research/sota/audio-visual-video-parsing-on-llp-val.md?metric=Event-Level%20Visual%20Accuracy>): Higher is better; unit: PERCENT.

- [Event-Level Audio-Visual Accuracy](<https://www.sota2.com/research/sota/audio-visual-video-parsing-on-llp-val.md?metric=Event-Level%20Audio-Visual%20Accuracy>): Higher is better; unit: PERCENT.

- [Event-Level Type Accuracy](<https://www.sota2.com/research/sota/audio-visual-video-parsing-on-llp-val.md?metric=Event-Level%20Type%20Accuracy>): Higher is better; unit: PERCENT.

- [Event-Level Event Accuracy](<https://www.sota2.com/research/sota/audio-visual-video-parsing-on-llp-val.md?metric=Event-Level%20Event%20Accuracy>): Higher is better; unit: PERCENT.

- [Average Accuracy](<https://www.sota2.com/research/sota/audio-visual-video-parsing-on-llp-val.md?metric=Average%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Segment-Level Audio Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| EAR | 83.4 | Not recorded | 6a02b437dbc836b87d824854 |
| VALOR | 80.8 | Not recorded | 6a02b437dbc836b87d824854 |
| UWAV | 80.1 | Not recorded | 6a02b437dbc836b87d824854 |
| BL+RegBN | 63.5 | Not recorded | 68fa71411e209131cea502c6 |
| BL | 61.8 | Not recorded | 68fa71411e209131cea502c6 |
| BL+PMDN | 61.4 | Not recorded | 68fa71411e209131cea502c6 |

### EAR

Segment-Level Audio Accuracy: 83.4. Result ID: 6a02f33647f34ce924e604a9.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b437dbc836b87d824854.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a02b437dbc836b87d824854.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### VALOR

Segment-Level Audio Accuracy: 80.8. Result ID: 6a02f33647f34ce924e604a7.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b437dbc836b87d824854.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70bd1e209131cea4cfa6.md>)

[Implementation](<https://github.com/Franklin905/VALOR>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### UWAV

Segment-Level Audio Accuracy: 80.1. Result ID: 6a02f33647f34ce924e604a8.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b437dbc836b87d824854.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78841e209131cea78c94.md>)

[Implementation](<https://www.sota2.com/Code%20Link>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### BL+RegBN

Segment-Level Audio Accuracy: 63.5. Result ID: 6991c1c93a6c3148490a60a7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71411e209131cea502c6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71411e209131cea502c6.md>)

[Implementation](<https://mogvision.github.io/RegBN>)

Recorded evaluation configuration: {"Backbone":"AVVP","Normalization":"RegBN","Learnable parameters":"4,571,723"}. This record does not assert independent reproduction.

### BL

Segment-Level Audio Accuracy: 61.8. Result ID: 6991c1c93a6c3148490a60a5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71411e209131cea502c6.md>)

Recorded evaluation configuration: {"Backbone":"AVVP","Learnable parameters":"4,571,723"}. This record does not assert independent reproduction.

### BL+PMDN

Segment-Level Audio Accuracy: 61.4. Result ID: 6991c1c93a6c3148490a60a6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71411e209131cea502c6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cf51e209131cea36dc7.md>)

[Implementation](<https://github.com/vento99/PMDN>)

Recorded evaluation configuration: {"Backbone":"AVVP","Normalization":"PMDN","Learnable parameters":"30,786,123"}. This record does not assert independent reproduction.

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