STPG (8+16frames)

Reported on 4 benchmarks across 2 tasks

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Robots2 results

Activity RecognitiononSomething-Something V1
Top 1 Accuracy
53.5
best: 70 (InternVideo)
Activity RecognitiononSomething-Something V2
Top-1 Accuracy
67
best: 77.3 (MVD (Kinetics400 pretrain, ViT-H, 16 frame))

Time Series2 results

Action RecognitiononSomething-Something V1
Top 1 Accuracy
53.5
best: 70 (InternVideo)
Action RecognitiononSomething-Something V2
Top-1 Accuracy
67
best: 77.3 (MVD (Kinetics400 pretrain, ViT-H, 16 frame))