# Action Triplet Recognition on Surgical Video Dataset

Source: [Action Triplet Recognition on Surgical Video Dataset](<https://www.sota2.com/research/sota/action-triplet-recognition-on-surgical-video-dataset>)

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

Data updated: 2026-02-26T05:23:14.598Z

Evaluation of action triplet recognition performance on the Surgical Video Dataset, measured by metrics such as AP-I, AP-V, AP-T, AP-IV, AP-IT, and AP-IVT.

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 Triplet Recognition |
| Dataset (including any stated split) | Surgical Video Dataset |
| Metric | AP-I |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| 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

- [AP-I](<https://www.sota2.com/research/sota/action-triplet-recognition-on-surgical-video-dataset.md?metric=AP-I>): Higher is better; unit: PERCENT.

- [AP-V](<https://www.sota2.com/research/sota/action-triplet-recognition-on-surgical-video-dataset.md?metric=AP-V>): Higher is better; unit: PERCENT.

- [AP-T](<https://www.sota2.com/research/sota/action-triplet-recognition-on-surgical-video-dataset.md?metric=AP-T>): Higher is better; unit: PERCENT.

- [AP-IV](<https://www.sota2.com/research/sota/action-triplet-recognition-on-surgical-video-dataset.md?metric=AP-IV>): Higher is better; unit: PERCENT.

- [AP-IT](<https://www.sota2.com/research/sota/action-triplet-recognition-on-surgical-video-dataset.md?metric=AP-IT>): Higher is better; unit: PERCENT.

- [AP-IVT](<https://www.sota2.com/research/sota/action-triplet-recognition-on-surgical-video-dataset.md?metric=AP-IVT>): Higher is better; unit: PERCENT.

## Recorded results

| Method | AP-I | Result date | Report paper |
| --- | --- | --- | --- |
| MML-SurgAdapt | 87.3 | Not recorded | 68fa79ed1e209131cea8001f |
| CLIPVIT-L/14-TS | 86.5 | Not recorded | 68fa79ed1e209131cea8001f |
| Rendezvous | 82 | Not recorded | 68fa79ed1e209131cea8001f |
| HSPNet | 77.6 | Not recorded | 68fa79ed1e209131cea8001f |
| VLPL | 69.7 | Not recorded | 68fa79ed1e209131cea8001f |
| ResNet50 TS | 64 | Not recorded | 68fa79ed1e209131cea8001f |
| Multi-task ResNet50 | 62 | Not recorded | 68fa79ed1e209131cea8001f |
| DualCoOp | 43.6 | Not recorded | 68fa79ed1e209131cea8001f |

### MML-SurgAdapt

AP-I: 87.3. Result ID: 698b65a07a967eddb76e3ed3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ed1e209131cea8001f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79ed1e209131cea8001f.md>)

[Implementation](<https://github.com/CAMMA-public/MML-SurgAdapt>)

Recorded evaluation configuration: {"Model Category":"SPML Models"}. This record does not assert independent reproduction.

### CLIPVIT-L/14-TS

AP-I: 86.5. Result ID: 698b65a07a967eddb76e3ece.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ed1e209131cea8001f.md>)

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

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

Recorded evaluation configuration: {"Model Category":"Task-Specific Models"}. This record does not assert independent reproduction.

### Rendezvous

AP-I: 82. Result ID: 698b65a07a967eddb76e3ecc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ed1e209131cea8001f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c301e209131cea32b2c.md>)

[Implementation](<https://github.com/CAMMA-public/rendezvous>)

Recorded evaluation configuration: {"Model Category":"Vision-only Models"}. This record does not assert independent reproduction.

### HSPNet

AP-I: 77.6. Result ID: 698b65a07a967eddb76e3ed2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ed1e209131cea8001f.md>)

Recorded evaluation configuration: {"Model Category":"SPML Models"}. This record does not assert independent reproduction.

### VLPL

AP-I: 69.7. Result ID: 698b65a07a967eddb76e3ed1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ed1e209131cea8001f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70f11e209131cea4e359.md>)

Recorded evaluation configuration: {"Model Category":"SPML Models"}. This record does not assert independent reproduction.

### ResNet50 TS

AP-I: 64. Result ID: 698b65a07a967eddb76e3ecd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ed1e209131cea8001f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62321e209131ce9f8328.md>)

Recorded evaluation configuration: {"Model Category":"Task-Specific Models"}. This record does not assert independent reproduction.

### Multi-task ResNet50

AP-I: 62. Result ID: 698b65a07a967eddb76e3ecf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ed1e209131cea8001f.md>)

Recorded evaluation configuration: {"Model Category":"Task-Specific Models"}. This record does not assert independent reproduction.

### DualCoOp

AP-I: 43.6. Result ID: 698b65a07a967eddb76e3ed0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79ed1e209131cea8001f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cdb1e209131cea3653f.md>)

Recorded evaluation configuration: {"Model Category":"SPML Models"}. This record does not assert independent reproduction.

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