Tasks SotA Datasets Papers Methods Submit About

Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable Benchmarks All SotA Datasets Papers Methods

Community

Submit Results About

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Models/VIMPAC

VIMPAC

Reported on 9 benchmarks across 3 tasks · 1 paper · 2 SOTA

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

Robots4 results

Activity RecognitiononDiving-48
Accuracy· 2021-06-21
85.5
best: 94.9 (LVMAE)
SOTA
VIMPAC: Video Pre-Training via Masked Token Prediction and Contrastive Learning arXiv:2106.11250
Activity RecognitiononHMDB-51
Average accuracy of 3 splits· 2021-06-21
65.9
best: 88.7 (VideoMAE V2-g)
VIMPAC: Video Pre-Training via Masked Token Prediction and Contrastive Learning arXiv:2106.11250
Activity RecognitiononSomething-Something V2
Top-1 Accuracy· uses extra data· 2021-06-21
68.1
best: 77.3 (MVD (Kinetics400 pretrain, ViT-H, 16 frame))
VIMPAC: Video Pre-Training via Masked Token Prediction and Contrastive Learning arXiv:2106.11250
Activity RecognitiononUCF101
3-fold Accuracy· 2021-06-21
92.7
best: 99.7 (FTP-UniFormerV2-L/14)
VIMPAC: Video Pre-Training via Masked Token Prediction and Contrastive Learning arXiv:2106.11250

Time Series4 results

Action RecognitiononDiving-48
Accuracy· 2021-06-21
85.5
best: 94.9 (LVMAE)
SOTA
VIMPAC: Video Pre-Training via Masked Token Prediction and Contrastive Learning arXiv:2106.11250
Action RecognitiononHMDB-51
Average accuracy of 3 splits· 2021-06-21
65.9
best: 88.7 (VideoMAE V2-g)
VIMPAC: Video Pre-Training via Masked Token Prediction and Contrastive Learning arXiv:2106.11250
Action RecognitiononSomething-Something V2
Top-1 Accuracy· uses extra data· 2021-06-21
68.1
best: 77.3 (MVD (Kinetics400 pretrain, ViT-H, 16 frame))
VIMPAC: Video Pre-Training via Masked Token Prediction and Contrastive Learning arXiv:2106.11250
Action RecognitiononUCF101
3-fold Accuracy· 2021-06-21
92.7
best: 99.7 (FTP-UniFormerV2-L/14)
VIMPAC: Video Pre-Training via Masked Token Prediction and Contrastive Learning arXiv:2106.11250

Computer Vision1 result

VideoonKinetics-400
Acc@1· 2021-06-21
77.4
best: 93.6 (OmniVec2)
VIMPAC: Video Pre-Training via Masked Token Prediction and Contrastive Learning arXiv:2106.11250