# Video Question Answering on ETBench

Source: [Video Question Answering on ETBench](<https://www.sota2.com/research/sota/video-question-answering-on-etbench>)

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

Data updated: 2026-02-26T06:06:03.159Z

Evaluation of model performance on the ETBench dataset for Video Question Answering capabilities, reporting specific component metrics like TVG F1, EPM F1, and DVC 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 | Video Question Answering |
| Dataset (including any stated split) | ETBench |
| Metric | TVG F1 |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 13 of 13 |
| 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

- [TVG F1](<https://www.sota2.com/research/sota/video-question-answering-on-etbench.md?metric=TVG%20F1>): Higher is better; unit: PERCENT.

- [EPM F1](<https://www.sota2.com/research/sota/video-question-answering-on-etbench.md?metric=EPM%20F1>): Higher is better; unit: PERCENT.

- [TAL F1](<https://www.sota2.com/research/sota/video-question-answering-on-etbench.md?metric=TAL%20F1>): Higher is better; unit: PERCENT.

- [VHD F1](<https://www.sota2.com/research/sota/video-question-answering-on-etbench.md?metric=VHD%20F1>): Higher is better; unit: PERCENT.

- [DVC F1](<https://www.sota2.com/research/sota/video-question-answering-on-etbench.md?metric=DVC%20F1>): Higher is better; unit: PERCENT.

- [DVC Similarity](<https://www.sota2.com/research/sota/video-question-answering-on-etbench.md?metric=DVC%20Similarity>): Higher is better; unit: PERCENT.

- [SLC F1](<https://www.sota2.com/research/sota/video-question-answering-on-etbench.md?metric=SLC%20F1>): Higher is better; unit: PERCENT.

- [SLC Similarity](<https://www.sota2.com/research/sota/video-question-answering-on-etbench.md?metric=SLC%20Similarity>): Higher is better; unit: PERCENT.

## Recorded results

| Method | TVG F1 | Result date | Report paper |
| --- | --- | --- | --- |
| Dispider | 43.6 | Not recorded | 68fa767f1e209131cea6d744 |
| ETChat | 38.6 | Not recorded | 68fa767f1e209131cea6d744 |
| Dispider | 36.1 | Not recorded | 68fa767f1e209131cea6d744 |
| TimeChat | 26.2 | Not recorded | 68fa767f1e209131cea6d744 |
| LITA | 22.2 | Not recorded | 68fa767f1e209131cea6d744 |
| VTG-LLM | 15.9 | Not recorded | 68fa767f1e209131cea6d744 |
| VideoLLM-Online | 13.2 | Not recorded | 68fa767f1e209131cea6d744 |
| VTimeLLM | 7.6 | Not recorded | 68fa767f1e209131cea6d744 |
| VideoChatGPT | 7 | Not recorded | 68fa767f1e209131cea6d744 |
| Video-LLaVA | 7 | Not recorded | 68fa767f1e209131cea6d744 |
| PLLAVA | 6.9 | Not recorded | 68fa767f1e209131cea6d744 |
| LLaMA-VID | 5.5 | Not recorded | 68fa767f1e209131cea6d744 |
| Video-LLaMA2 | 0.1 | Not recorded | 68fa767f1e209131cea6d744 |

### Dispider

TVG F1: 43.6. Result ID: 69943ee23a6c3148491c0809.

[Reporting paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Implementation](<https://github.com/Mark12Ding/Dispider>)

Recorded evaluation configuration: {"Frames":"1 FPS","Inference Setting":"Conventional video QA inference","Specialized Time Tokens":false}. This record does not assert independent reproduction.

### ETChat

TVG F1: 38.6. Result ID: 69943ee23a6c3148491c0801.

[Reporting paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74f31e209131cea64c2c.md>)

[Implementation](<https://polyu-chenlab.github.io/etbench/>)

Recorded evaluation configuration: {"Frames":"1 FPS","Inference Setting":"Conventional video QA inference","Specialized Time Tokens":true}. This record does not assert independent reproduction.

### Dispider

TVG F1: 36.1. Result ID: 69943ee23a6c3148491c080b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Implementation](<https://github.com/Mark12Ding/Dispider>)

Recorded evaluation configuration: {"Frames":"1 FPS","Inference Setting":"Streaming video QA inference"}. This record does not assert independent reproduction.

### TimeChat

TVG F1: 26.2. Result ID: 69943ee23a6c3148491c0808.

[Reporting paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72ac1e209131cea57c30.md>)

[Implementation](<https://github.com/RenShuhuai-Andy/TimeChat>)

Recorded evaluation configuration: {"Frames":"96","Inference Setting":"Conventional video QA inference","Specialized Time Tokens":false}. This record does not assert independent reproduction.

### LITA

TVG F1: 22.2. Result ID: 69943ee23a6c3148491c0800.

[Reporting paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72ad1e209131cea57cb4.md>)

[Implementation](<https://github.com/NVlabs/LITA>)

Recorded evaluation configuration: {"Frames":"100","Inference Setting":"Conventional video QA inference","Specialized Time Tokens":true}. This record does not assert independent reproduction.

### VTG-LLM

TVG F1: 15.9. Result ID: 69943ee23a6c3148491c07ff.

[Reporting paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76c61e209131cea6eefa.md>)

Recorded evaluation configuration: {"Frames":"96","Inference Setting":"Conventional video QA inference","Specialized Time Tokens":true}. This record does not assert independent reproduction.

### VideoLLM-Online

TVG F1: 13.2. Result ID: 69943ee23a6c3148491c080a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73a71e209131cea5db14.md>)

[Implementation](<https://www.sota2.com/showlab.github.io/videollm-online>)

Recorded evaluation configuration: {"Frames":"2 FPS","Inference Setting":"Streaming video QA inference"}. This record does not assert independent reproduction.

### VTimeLLM

TVG F1: 7.6. Result ID: 69943ee23a6c3148491c0807.

[Reporting paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa715b1e209131cea50d56.md>)

[Implementation](<https://github.com/huangb23/VTimeLLM>)

Recorded evaluation configuration: {"Frames":"100","Inference Setting":"Conventional video QA inference","Specialized Time Tokens":false}. This record does not assert independent reproduction.

### VideoChatGPT

TVG F1: 7. Result ID: 69943ee23a6c3148491c0802.

[Reporting paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73841e209131cea5ce32.md>)

[Implementation](<https://github.com/mbzuai-oryx/Video-ChatGPT>)

Recorded evaluation configuration: {"Frames":"100","Inference Setting":"Conventional video QA inference","Specialized Time Tokens":false}. This record does not assert independent reproduction.

### Video-LLaVA

TVG F1: 7. Result ID: 69943ee23a6c3148491c0803.

[Reporting paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75051e209131cea65534.md>)

[Implementation](<https://github.com/PKU-YuanGroup/Video-LLaVA>)

Recorded evaluation configuration: {"Frames":"8","Inference Setting":"Conventional video QA inference","Specialized Time Tokens":false}. This record does not assert independent reproduction.

### PLLAVA

TVG F1: 6.9. Result ID: 69943ee23a6c3148491c0806.

[Reporting paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73101e209131cea5a3ea.md>)

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

Recorded evaluation configuration: {"Frames":"16","Inference Setting":"Conventional video QA inference","Specialized Time Tokens":false}. This record does not assert independent reproduction.

### LLaMA-VID

TVG F1: 5.5. Result ID: 69943ee23a6c3148491c0804.

[Reporting paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71461e209131cea50559.md>)

[Implementation](<https://github.com/dvlab-research/LLaMA-VID>)

Recorded evaluation configuration: {"Frames":"1 FPS","Inference Setting":"Conventional video QA inference","Specialized Time Tokens":false}. This record does not assert independent reproduction.

### Video-LLaMA2

TVG F1: 0.1. Result ID: 69943ee23a6c3148491c0805.

[Reporting paper](<https://www.sota2.com/research/paper/68fa767f1e209131cea6d744.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75761e209131cea681dd.md>)

[Implementation](<https://github.com/DAMO-NLP-SG/VideoLLaMA2>)

Recorded evaluation configuration: {"Frames":"8","Inference Setting":"Conventional video QA inference","Specialized Time Tokens":false}. This record does not assert independent reproduction.

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