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/VideoMAE V2-g

VideoMAE V2-g

Reported on 37 benchmarks across 6 tasks · 1 paper · 29 SOTA

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

Computer Vision17 results

VideoonFineAction
mAP· uses extra data· 2023-03-29
18.24
best: 29.6 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
VideoonFineAction
mAP IOU@0.5· uses extra data· 2023-03-29
29.07
best: 46.4 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
VideoonFineAction
mAP IOU@0.75· uses extra data· 2023-03-29
17.66
best: 29.5 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
VideoonFineAction
mAP IOU@0.95· uses extra data· 2023-03-29
5.07
best: 7.6 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Temporal Action LocalizationonFineAction
mAP· uses extra data· 2023-03-29
18.24
best: 29.6 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Temporal Action LocalizationonFineAction
mAP IOU@0.5· uses extra data· 2023-03-29
29.07
best: 46.4 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Temporal Action LocalizationonFineAction
mAP IOU@0.75· uses extra data· 2023-03-29
17.66
best: 29.5 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Temporal Action LocalizationonFineAction
mAP IOU@0.95· uses extra data· 2023-03-29
5.07
best: 7.6 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Action LocalizationonFineAction
mAP· uses extra data· 2023-03-29
18.24
best: 29.6 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Action LocalizationonFineAction
mAP IOU@0.5· uses extra data· 2023-03-29
29.07
best: 46.4 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Action LocalizationonFineAction
mAP IOU@0.75· uses extra data· 2023-03-29
17.66
best: 29.5 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Action LocalizationonFineAction
mAP IOU@0.95· uses extra data· 2023-03-29
5.07
best: 7.6 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Action LocalizationonAVA-Kinetics
val mAP· uses extra data· 2023-03-29
42.6
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
VideoonKinetics-400
Acc@1· uses extra data· 2023-03-29
88.5
best: 93.6 (OmniVec2)
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
VideoonKinetics-400
Acc@5· uses extra data· 2023-03-29
98.1
best: 98.9 (TubeViT-H (ImageNet-1k))
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
VideoonKinetics-600
Top-1 Accuracy· uses extra data· 2023-03-29
88.8
best: 91.9 (InternVideo2-6B)
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
VideoonKinetics-600
Top-5 Accuracy· uses extra data· 2023-03-29
98.2
best: 98.9 (TubeVit-H)
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727

Robots9 results

Activity RecognitiononHMDB-51
Average accuracy of 3 splits· uses extra data· 2023-03-29
88.7
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Activity RecognitiononSomething-Something V1
Top 5 Accuracy· uses extra data· 2023-03-29
91.9
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Activity RecognitiononSomething-Something V2
Parameters· uses extra data· 2023-03-29
1013
best: 2131 (InternVideo2-6B)
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Activity RecognitiononSomething-Something V2
Top-5 Accuracy· uses extra data· 2023-03-29
95.9
best: 96.3 (DejaVid)
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Activity RecognitiononUCF101
3-fold Accuracy· uses extra data· 2023-03-29
99.6
best: 99.7 (FTP-UniFormerV2-L/14)
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Activity RecognitiononAVA v2.2
mAP· uses extra data· 2023-03-29
42.6
best: 45.1 (LART (Hiera-H, K700 PT+FT))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Activity RecognitiononSomething-Something V1
Top 1 Accuracy· uses extra data· 2023-03-29
68.7
best: 70 (InternVideo)
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Activity RecognitiononSomething-Something V2
Top-1 Accuracy· uses extra data· 2023-03-29
77
best: 77.3 (MVD (Kinetics400 pretrain, ViT-H, 16 frame))
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Activity RecognitiononUCF101
3-fold Accuracy· uses extra data· 2023-03-29
99.6
best: 99.7 (FTP-UniFormerV2-L/14)
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727

Time Series9 results

Action RecognitiononHMDB-51
Average accuracy of 3 splits· uses extra data· 2023-03-29
88.7
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Action RecognitiononSomething-Something V1
Top 5 Accuracy· uses extra data· 2023-03-29
91.9
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Action RecognitiononSomething-Something V2
Parameters· uses extra data· 2023-03-29
1013
best: 2131 (InternVideo2-6B)
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Action RecognitiononSomething-Something V2
Top-5 Accuracy· uses extra data· 2023-03-29
95.9
best: 96.3 (DejaVid)
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Action RecognitiononUCF101
3-fold Accuracy· uses extra data· 2023-03-29
99.6
best: 99.7 (FTP-UniFormerV2-L/14)
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Action RecognitiononAVA v2.2
mAP· uses extra data· 2023-03-29
42.6
best: 45.1 (LART (Hiera-H, K700 PT+FT))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Action RecognitiononSomething-Something V1
Top 1 Accuracy· uses extra data· 2023-03-29
68.7
best: 70 (InternVideo)
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Action RecognitiononSomething-Something V2
Top-1 Accuracy· uses extra data· 2023-03-29
77
best: 77.3 (MVD (Kinetics400 pretrain, ViT-H, 16 frame))
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Action RecognitiononUCF101
3-fold Accuracy· uses extra data· 2023-03-29
99.6
best: 99.7 (FTP-UniFormerV2-L/14)
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727

Methodology4 results

Zero-Shot LearningonFineAction
mAP· uses extra data· 2023-03-29
18.24
best: 29.6 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Zero-Shot LearningonFineAction
mAP IOU@0.5· uses extra data· 2023-03-29
29.07
best: 46.4 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Zero-Shot LearningonFineAction
mAP IOU@0.75· uses extra data· 2023-03-29
17.66
best: 29.5 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727
Zero-Shot LearningonFineAction
mAP IOU@0.95· uses extra data· 2023-03-29
5.07
best: 7.6 (RDFA-S6 (InternVideo2-6B))
SOTA
VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking arXiv:2303.16727