# Video Question Answering on EgoSchema ES-S (public subset)

Source: [Video Question Answering on EgoSchema ES-S (public subset)](<https://www.sota2.com/research/sota/video-question-answering-on-egoschema-es-s-public-subset>)

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

Data updated: 2026-02-26T05:30:38.772Z

Evaluation of model performance using Accuracy and Inference Time on the public subset (ES-S) of the EgoSchema dataset for Video Question Answering.

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 Question Answering |
| Dataset (including any stated split) | EgoSchema ES-S (public subset) |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher 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

- [Accuracy](<https://www.sota2.com/research/sota/video-question-answering-on-egoschema-es-s-public-subset.md?metric=Accuracy>): Higher is better; unit: PERCENT.

- [Inference Time per Sample](<https://www.sota2.com/research/sota/video-question-answering-on-egoschema-es-s-public-subset.md?metric=Inference%20Time%20per%20Sample>): Lower is better; unit: TIME.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| SF-VLM | 55.8 | Not recorded | 68fa795c1e209131cea7d42d |
| SOTA (Zhang et al. 2023a) | 50.8 | Not recorded | 68fa795c1e209131cea7d42d |
| Just-LLM | 45.8 | Not recorded | 68fa795c1e209131cea7d42d |
| Random | 20 | Not recorded | 68fa795c1e209131cea7d42d |

### SF-VLM

Accuracy: 55.8. Result ID: 6995dc713badfaffe6e31fac.

[Reporting paper](<https://www.sota2.com/research/paper/68fa795c1e209131cea7d42d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa795c1e209131cea7d42d.md>)

[Implementation](<https://github.com/kahnchana/mvu>)

Recorded evaluation configuration: {"Param":"13B","Video Frames":1}. This record does not assert independent reproduction.

### SOTA (Zhang et al. 2023a)

Accuracy: 50.8. Result ID: 6995dc713badfaffe6e31fad.

[Reporting paper](<https://www.sota2.com/research/paper/68fa795c1e209131cea7d42d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75261e209131cea6629c.md>)

[Implementation](<https://github.com/CeeZh/LLoVi>)

Recorded evaluation configuration: {"Param":"20B","Video Frames":180}. This record does not assert independent reproduction.

### Just-LLM

Accuracy: 45.8. Result ID: 6995dc713badfaffe6e31fab.

[Reporting paper](<https://www.sota2.com/research/paper/68fa795c1e209131cea7d42d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa795c1e209131cea7d42d.md>)

[Implementation](<https://github.com/kahnchana/mvu>)

Recorded evaluation configuration: {"Param":"7B","Video Frames":0}. This record does not assert independent reproduction.

### Random

Accuracy: 20. Result ID: 6995dc713badfaffe6e31faa.

[Reporting paper](<https://www.sota2.com/research/paper/68fa795c1e209131cea7d42d.md>)

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

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