List your company
ResearchFormerly WizWand

Action Recognition on Something-Something V2 (Base, Novel, HM)

19.2Base ScoreVideo-STAR-7B
2.8727.11111.3515.589Dec 6, 2022May 28, 2023Nov 17, 2023May 8, 2024Oct 28, 2024Apr 19, 2025Oct 9, 2025
Updated 6mo ago

Evaluation results

MethodLinks
Video-STAR-7BBackbone=ViT-B/16, Evaluation Setting=base-to-novel2025.1019.21315.5
FROSTERBackbone=ViT-B/16, Evaluation Setting=base-to-novel2025.1018.312.214.6
VTD-CLIPBackbone=ViT-B/16, Evaluation Setting=base-to-novel2025.1017.813.915.4
Open-MeDeBackbone=ViT-B/16, Evaluation Setting=base-to-novel2025.1017.112.314.3
EZ-CLIPbackbone=ViT-162023.1216.613.314.8
AP-CLIPBackbone=ViT-B/16, Evaluation Setting=base-to-novel2025.1016.312.914.4
ViFi-CLIPAdaptation Protocol=Tuning pre-trained image VL models, Backbone=ViT-B/16, Number of frames=322022.1216.212.113.9
ViFi CLIPSource=Rasheed et al. (2023)2023.1216.212.113.9
ViFi-CLIPBackbone=ViT-B/16, Evaluation Setting=base-to-novel2025.1016.212.113.9
Qwen2.5-VL-7BBackbone=ViT-B/16, Evaluation Setting=base-to-novel2025.10149.911.6
Video-STAR-3BBackbone=ViT-B/16, Evaluation Setting=base-to-novel2025.1013.511.312.3
ActionCLIPAdaptation Protocol=Adapting pre-trained image VL models, Backbone=ViT-B/16, Number of frames=322022.1213.310.111.5
ActionCLIPSource=Wang et al. (2021)2023.1213.310.111.5
ActionCLIPBackbone=ViT-B/16, Evaluation Setting=base-to-novel2025.1013.310.111.5
CLIP text-FTAdaptation Protocol=Tuning pre-trained image VL models, Backbone=ViT-B/16, Number of frames=322022.1212.49.510.8
ST-AdapterBackbone=ViT-B/16, Evaluation Setting=base-to-novel2025.109.38.48.8
CLIP image-FTAdaptation Protocol=Tuning pre-trained image VL models, Backbone=ViT-B/16, Number of frames=322022.129.28.58.8
XCLIPAdaptation Protocol=Adapting pre-trained image VL models, Backbone=ViT-B/16, Number of frames=322022.128.56.67.4
XCLIPSource=Ni et al. (2022)2023.128.56.67.4
X-CLIPBackbone=ViT-B/16, Evaluation Setting=base-to-novel2025.108.56.67.4
A5Adaptation Protocol=Adapting pre-trained image VL models, Backbone=ViT-B/16, Number of frames=322022.128.35.36.4
A5Source=Ju et al. (2022)2023.128.35.36.4
VPTBackbone=ViT-B/16, Evaluation Setting=base-to-novel2025.108.35.36.4
Vanilla CLIPAdaptation Protocol=Adapting pre-trained image VL models, Backbone=ViT-B/16, Number of frames=322022.124.95.35.1
Vanila CLIPSource=Radford et al. (2021)2023.124.95.35.1
Qwen2.5-VL-3BBackbone=ViT-B/16, Evaluation Setting=base-to-novel2025.103.53.23.3