# Audio-Visual Scene-aware Dialog on AVSD (val)

Source: [Audio-Visual Scene-aware Dialog on AVSD (val)](<https://www.sota2.com/research/sota/audio-visual-scene-aware-dialog-on-avsd-val>)

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

Data updated: 2026-02-26T05:24:00.720Z

Evaluation of Audio-Visual Scene-aware Dialog performance on the validation split of the AVSD dataset, measured using ASR percentage.

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 Scene-aware Dialog |
| Dataset (including any stated split) | AVSD (val) |
| Metric | ASR (%) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 7 of 7 |
| 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

- [ASR (%)](<https://www.sota2.com/research/sota/audio-visual-scene-aware-dialog-on-avsd-val.md?metric=ASR%20(%25)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | ASR (%) | Result date | Report paper |
| --- | --- | --- | --- |
| Combined Loss | 59.48 | Not recorded | 6976f519a2b28e761d87c22c |
| Encoder-Based Cosine Similarity Loss | 57.03 | Not recorded | 6976f519a2b28e761d87c22c |
| Negative Language Modeling Loss | 50.43 | Not recorded | 6976f519a2b28e761d87c22c |
| Vision Attention Suppression Loss | 49.34 | Not recorded | 6976f519a2b28e761d87c22c |
| Attention Randomization Loss | 45.67 | Not recorded | 6976f519a2b28e761d87c22c |
| Audio Attention Amplification Loss | 43.85 | Not recorded | 6976f519a2b28e761d87c22c |
| Hidden-State Similarity Loss | 41.53 | Not recorded | 6976f519a2b28e761d87c22c |

### Combined Loss

ASR (%): 59.48. Result ID: 699a052b5698665ec4878b97.

[Reporting paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

Recorded evaluation configuration: {"Attack Objective":"L(combined)","Source Model":"VideoLLAMA2","Training Dataset":"AVQA","Training Epochs":"150"}. This record does not assert independent reproduction.

### Encoder-Based Cosine Similarity Loss

ASR (%): 57.03. Result ID: 699a052b5698665ec4878b92.

[Reporting paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

Recorded evaluation configuration: {"Attack Objective":"L(cos)","Source Model":"VideoLLAMA2","Training Dataset":"AVQA","Training Epochs":"150"}. This record does not assert independent reproduction.

### Negative Language Modeling Loss

ASR (%): 50.43. Result ID: 699a052b5698665ec4878b91.

[Reporting paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

Recorded evaluation configuration: {"Attack Objective":"L\_negLM","Source Model":"VideoLLAMA2","Training Dataset":"AVQA","Training Epochs":"150"}. This record does not assert independent reproduction.

### Vision Attention Suppression Loss

ASR (%): 49.34. Result ID: 699a052b5698665ec4878b93.

[Reporting paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

Recorded evaluation configuration: {"Attack Objective":"L(visionatt)","Source Model":"VideoLLAMA2","Training Dataset":"AVQA","Training Epochs":"150"}. This record does not assert independent reproduction.

### Attention Randomization Loss

ASR (%): 45.67. Result ID: 699a052b5698665ec4878b95.

[Reporting paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

Recorded evaluation configuration: {"Attack Objective":"L(randatt)","Source Model":"VideoLLAMA2","Training Dataset":"AVQA","Training Epochs":"150"}. This record does not assert independent reproduction.

### Audio Attention Amplification Loss

ASR (%): 43.85. Result ID: 699a052b5698665ec4878b94.

[Reporting paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

Recorded evaluation configuration: {"Attack Objective":"L(audioatt)","Source Model":"VideoLLAMA2","Training Dataset":"AVQA","Training Epochs":"150"}. This record does not assert independent reproduction.

### Hidden-State Similarity Loss

ASR (%): 41.53. Result ID: 699a052b5698665ec4878b96.

[Reporting paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6976f519a2b28e761d87c22c.md>)

Recorded evaluation configuration: {"Attack Objective":"L(hidden-cos)","Source Model":"VideoLLAMA2","Training Dataset":"AVQA","Training Epochs":"150"}. This record does not assert independent reproduction.

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