# Video Classification on Kinetics-400 and Something-Something V2

Source: [Video Classification on Kinetics-400 and Something-Something V2](<https://www.sota2.com/research/sota/video-classification-on-kinetics-400-and-something-something-v2>)

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

Data updated: 2026-02-26T05:47:41.574Z

Evaluation of video classification performance across both the Kinetics-400 and Something-Something V2 datasets, reporting individual and average results.

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 Classification |
| Dataset (including any stated split) | Video Datasets K400, SSv2 |
| Metric | K400 Accuracy |
| 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

- [K400 Accuracy](<https://www.sota2.com/research/sota/video-classification-on-kinetics-400-and-something-something-v2.md?metric=K400%20Accuracy>): Higher is better; unit: PERCENT.

- [SSv2 Accuracy](<https://www.sota2.com/research/sota/video-classification-on-kinetics-400-and-something-something-v2.md?metric=SSv2%20Accuracy>): Higher is better; unit: PERCENT.

- [Average Accuracy](<https://www.sota2.com/research/sota/video-classification-on-kinetics-400-and-something-something-v2.md?metric=Average%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | K400 Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| AdaptFormer | 75.53 | Not recorded | 68fa753d1e209131cea66acc |
| Full tuning | 75.48 | Not recorded | 68fa753d1e209131cea66acc |
| LoRA | 75.48 | Not recorded | 68fa753d1e209131cea66acc |
| DyT + N = 4 | 75.43 | Not recorded | 68fa753d1e209131cea66acc |
| DyT | 75.28 | Not recorded | 68fa753d1e209131cea66acc |
| Dynamic-Full | 74.63 | Not recorded | 68fa753d1e209131cea66acc |
| VPT | 73.46 | Not recorded | 68fa753d1e209131cea66acc |
| Linear | 69.04 | Not recorded | 68fa753d1e209131cea66acc |

### AdaptFormer

K400 Accuracy: 75.53. Result ID: 699383533a6c314849167e74.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753d1e209131cea66acc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dab1e209131cea3b5b4.md>)

[Implementation](<https://github.com/ShoufaChen/AdaptFormer>)

Recorded evaluation configuration: {"Params (M)":"1.19","FLOPS (G)":"144.39"}. This record does not assert independent reproduction.

### Full tuning

K400 Accuracy: 75.48. Result ID: 699383533a6c314849167e71.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753d1e209131cea66acc.md>)

Recorded evaluation configuration: {"Params (M)":"85.80","FLOPS (G)":"142.53"}. This record does not assert independent reproduction.

### LoRA

K400 Accuracy: 75.48. Result ID: 699383533a6c314849167e75.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753d1e209131cea66acc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b121e209131cea2bf59.md>)

[Implementation](<https://github.com/microsoft/LoRA>)

Recorded evaluation configuration: {"Params (M)":"1.19","FLOPS (G)":"142.53"}. This record does not assert independent reproduction.

### DyT + N = 4

K400 Accuracy: 75.43. Result ID: 699383533a6c314849167e78.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753d1e209131cea66acc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa753d1e209131cea66acc.md>)

[Implementation](<https://github.com/NUS-HPC-AI-Lab/Dynamic-Tuning>)

Recorded evaluation configuration: {"Params (M)":"4.80","FLOPS (G)":"105.45"}. This record does not assert independent reproduction.

### DyT

K400 Accuracy: 75.28. Result ID: 699383533a6c314849167e77.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753d1e209131cea66acc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa753d1e209131cea66acc.md>)

[Implementation](<https://github.com/NUS-HPC-AI-Lab/Dynamic-Tuning>)

Recorded evaluation configuration: {"Params (M)":"1.19","FLOPS (G)":"108.31"}. This record does not assert independent reproduction.

### Dynamic-Full

K400 Accuracy: 74.63. Result ID: 699383533a6c314849167e73.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753d1e209131cea66acc.md>)

Recorded evaluation configuration: {"Params (M)":"85.80","FLOPS (G)":"107.67"}. This record does not assert independent reproduction.

### VPT

K400 Accuracy: 73.46. Result ID: 699383533a6c314849167e76.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753d1e209131cea66acc.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d181e209131cea37ae9.md>)

[Implementation](<https://www.sota2.com/github.com/kmnp/vpt>)

Recorded evaluation configuration: {"Params (M)":"0.07","FLOPS (G)":"148.44"}. This record does not assert independent reproduction.

### Linear

K400 Accuracy: 69.04. Result ID: 699383533a6c314849167e72.

[Reporting paper](<https://www.sota2.com/research/paper/68fa753d1e209131cea66acc.md>)

Recorded evaluation configuration: {"Params (M)":"0","FLOPS (G)":"142.53"}. This record does not assert independent reproduction.

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