# Action Recognition on HMDB-51 (Top-1 Accuracy)

Source: [Action Recognition on HMDB-51 (Top-1 Accuracy)](<https://www.sota2.com/research/sota/action-recognition-on-hmdb-51-top-1-accuracy>)

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

Data updated: 2026-05-25T08:19:05.867Z

Evaluation of video action recognition performance on the HMDB-51 dataset, reporting standard Top-1 Accuracy.

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 | Action Recognition |
| Dataset (including any stated split) | HMDB-51 |
| Metric | Top-1 Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 36 of 36 |
| 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

- [Top-1 Accuracy](<https://www.sota2.com/research/sota/action-recognition-on-hmdb-51-top-1-accuracy.md?metric=Top-1%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Top-1 Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| SimVA | 70.1 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| TC-CLIP | 68.8 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| OST | 68.2 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| BDC-CLIP | 67.5 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| SimVA | 67.5 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| ViFi-CLIP | 66.8 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| TC-CLIP | 65.5 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| BDC-CLIP | 65.3 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| OST | 64.9 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| ViFi-CLIP | 64.5 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| X-CLIP | 64 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| SimVA | 64 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| TC-CLIP | 63.3 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| BDC-CLIP | 63.1 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| OST | 62.9 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| X-CLIP | 62.8 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| ViFi-CLIP | 62.7 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| A5 | 62.4 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| SimVA | 60.7 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| BDC-CLIP | 59.2 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| ActionCLIP | 59.1 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| OST | 59.1 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| TC-CLIP | 58.6 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| ActionCLIP | 57.9 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| ActionCLIP | 57.3 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| X-CLIP | 57.3 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| ViFi-CLIP | 57.2 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| A5 | 56 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| X-CLIP | 53 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| A5 | 50.7 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| ActionCLIP | 47.5 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| CLIP | 41.9 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| CLIP | 41.9 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| CLIP | 41.9 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| CLIP | 41.9 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |
| A5 | 39.7 | Not recorded | 6a13d7586f71c5ae7dc9e6cb |

### SimVA

Top-1 Accuracy: 70.1. Result ID: 6a1405e2fbc373592c0b2077.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

Recorded evaluation configuration: {"K (Shots)":"16","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### TC-CLIP

Top-1 Accuracy: 68.8. Result ID: 6a1405e2fbc373592c0b206f.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74491e209131cea60c4a.md>)

[Implementation](<https://github.com/naver-ai/tc-clip>)

Recorded evaluation configuration: {"K (Shots)":"16","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### OST

Top-1 Accuracy: 68.2. Result ID: 6a1405e2fbc373592c0b206b.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72ac1e209131cea57c2f.md>)

[Implementation](<https://tomchen-ctj.github.io/OST>)

Recorded evaluation configuration: {"K (Shots)":"16","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### BDC-CLIP

Top-1 Accuracy: 67.5. Result ID: 6a1405e2fbc373592c0b2073.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

Recorded evaluation configuration: {"K (Shots)":"16","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### SimVA

Top-1 Accuracy: 67.5. Result ID: 6a1405e2fbc373592c0b2076.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

Recorded evaluation configuration: {"K (Shots)":"8","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### ViFi-CLIP

Top-1 Accuracy: 66.8. Result ID: 6a1405e2fbc373592c0b2067.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ee81e209131cea42cef.md>)

[Implementation](<https://github.com/muzairkhattak/ViFi-CLIP>)

Recorded evaluation configuration: {"K (Shots)":"16","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### TC-CLIP

Top-1 Accuracy: 65.5. Result ID: 6a1405e2fbc373592c0b206e.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74491e209131cea60c4a.md>)

[Implementation](<https://github.com/naver-ai/tc-clip>)

Recorded evaluation configuration: {"K (Shots)":"8","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### BDC-CLIP

Top-1 Accuracy: 65.3. Result ID: 6a1405e2fbc373592c0b2072.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

Recorded evaluation configuration: {"K (Shots)":"8","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### OST

Top-1 Accuracy: 64.9. Result ID: 6a1405e2fbc373592c0b206a.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72ac1e209131cea57c2f.md>)

[Implementation](<https://tomchen-ctj.github.io/OST>)

Recorded evaluation configuration: {"K (Shots)":"8","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### ViFi-CLIP

Top-1 Accuracy: 64.5. Result ID: 6a1405e2fbc373592c0b2066.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ee81e209131cea42cef.md>)

[Implementation](<https://github.com/muzairkhattak/ViFi-CLIP>)

Recorded evaluation configuration: {"K (Shots)":"8","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### X-CLIP

Top-1 Accuracy: 64. Result ID: 6a1405e2fbc373592c0b2063.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d701e209131cea39e2f.md>)

[Implementation](<https://github.com/xuguohai/X-CLIP>)

Recorded evaluation configuration: {"K (Shots)":"16","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### SimVA

Top-1 Accuracy: 64. Result ID: 6a1405e2fbc373592c0b2075.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

Recorded evaluation configuration: {"K (Shots)":"4","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### TC-CLIP

Top-1 Accuracy: 63.3. Result ID: 6a1405e2fbc373592c0b206d.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74491e209131cea60c4a.md>)

[Implementation](<https://github.com/naver-ai/tc-clip>)

Recorded evaluation configuration: {"K (Shots)":"4","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### BDC-CLIP

Top-1 Accuracy: 63.1. Result ID: 6a1405e2fbc373592c0b2071.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

Recorded evaluation configuration: {"K (Shots)":"4","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### OST

Top-1 Accuracy: 62.9. Result ID: 6a1405e2fbc373592c0b2069.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72ac1e209131cea57c2f.md>)

[Implementation](<https://tomchen-ctj.github.io/OST>)

Recorded evaluation configuration: {"K (Shots)":"4","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### X-CLIP

Top-1 Accuracy: 62.8. Result ID: 6a1405e2fbc373592c0b2062.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d701e209131cea39e2f.md>)

[Implementation](<https://github.com/xuguohai/X-CLIP>)

Recorded evaluation configuration: {"K (Shots)":"8","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### ViFi-CLIP

Top-1 Accuracy: 62.7. Result ID: 6a1405e2fbc373592c0b2065.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ee81e209131cea42cef.md>)

[Implementation](<https://github.com/muzairkhattak/ViFi-CLIP>)

Recorded evaluation configuration: {"K (Shots)":"4","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### A5

Top-1 Accuracy: 62.4. Result ID: 6a1405e2fbc373592c0b205b.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d041e209131cea375a1.md>)

Recorded evaluation configuration: {"K (Shots)":"16","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### SimVA

Top-1 Accuracy: 60.7. Result ID: 6a1405e2fbc373592c0b2074.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

Recorded evaluation configuration: {"K (Shots)":"2","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### BDC-CLIP

Top-1 Accuracy: 59.2. Result ID: 6a1405e2fbc373592c0b2070.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

Recorded evaluation configuration: {"K (Shots)":"2","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### ActionCLIP

Top-1 Accuracy: 59.1. Result ID: 6a1405e2fbc373592c0b205f.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6aca1e209131cea2a40f.md>)

[Implementation](<https://github.com/sallymmx/ActionCLIP.git>)

Recorded evaluation configuration: {"K (Shots)":"16","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### OST

Top-1 Accuracy: 59.1. Result ID: 6a1405e2fbc373592c0b2068.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72ac1e209131cea57c2f.md>)

[Implementation](<https://tomchen-ctj.github.io/OST>)

Recorded evaluation configuration: {"K (Shots)":"2","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### TC-CLIP

Top-1 Accuracy: 58.6. Result ID: 6a1405e2fbc373592c0b206c.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74491e209131cea60c4a.md>)

[Implementation](<https://github.com/naver-ai/tc-clip>)

Recorded evaluation configuration: {"K (Shots)":"2","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### ActionCLIP

Top-1 Accuracy: 57.9. Result ID: 6a1405e2fbc373592c0b205d.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6aca1e209131cea2a40f.md>)

[Implementation](<https://github.com/sallymmx/ActionCLIP.git>)

Recorded evaluation configuration: {"K (Shots)":"4","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### ActionCLIP

Top-1 Accuracy: 57.3. Result ID: 6a1405e2fbc373592c0b205e.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6aca1e209131cea2a40f.md>)

[Implementation](<https://github.com/sallymmx/ActionCLIP.git>)

Recorded evaluation configuration: {"K (Shots)":"8","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### X-CLIP

Top-1 Accuracy: 57.3. Result ID: 6a1405e2fbc373592c0b2061.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d701e209131cea39e2f.md>)

[Implementation](<https://github.com/xuguohai/X-CLIP>)

Recorded evaluation configuration: {"K (Shots)":"4","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### ViFi-CLIP

Top-1 Accuracy: 57.2. Result ID: 6a1405e2fbc373592c0b2064.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ee81e209131cea42cef.md>)

[Implementation](<https://github.com/muzairkhattak/ViFi-CLIP>)

Recorded evaluation configuration: {"K (Shots)":"2","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### A5

Top-1 Accuracy: 56. Result ID: 6a1405e2fbc373592c0b205a.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d041e209131cea375a1.md>)

Recorded evaluation configuration: {"K (Shots)":"8","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### X-CLIP

Top-1 Accuracy: 53. Result ID: 6a1405e2fbc373592c0b2060.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d701e209131cea39e2f.md>)

[Implementation](<https://github.com/xuguohai/X-CLIP>)

Recorded evaluation configuration: {"K (Shots)":"2","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### A5

Top-1 Accuracy: 50.7. Result ID: 6a1405e2fbc373592c0b2059.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d041e209131cea375a1.md>)

Recorded evaluation configuration: {"K (Shots)":"4","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### ActionCLIP

Top-1 Accuracy: 47.5. Result ID: 6a1405e2fbc373592c0b205c.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6aca1e209131cea2a40f.md>)

[Implementation](<https://github.com/sallymmx/ActionCLIP.git>)

Recorded evaluation configuration: {"K (Shots)":"2","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### CLIP

Top-1 Accuracy: 41.9. Result ID: 6a1405e2fbc373592c0b2054.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"K (Shots)":"2","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### CLIP

Top-1 Accuracy: 41.9. Result ID: 6a1405e2fbc373592c0b2055.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"K (Shots)":"4","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### CLIP

Top-1 Accuracy: 41.9. Result ID: 6a1405e2fbc373592c0b2056.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"K (Shots)":"8","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### CLIP

Top-1 Accuracy: 41.9. Result ID: 6a1405e2fbc373592c0b2057.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"K (Shots)":"16","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

### A5

Top-1 Accuracy: 39.7. Result ID: 6a1405e2fbc373592c0b2058.

[Reporting paper](<https://www.sota2.com/research/paper/6a13d7586f71c5ae7dc9e6cb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d041e209131cea375a1.md>)

Recorded evaluation configuration: {"K (Shots)":"2","Few-shot setting":"true","Fine-tuned from CLIP":"true"}. This record does not assert independent reproduction.

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