List your company
ResearchFormerly WizWand

Action Detection on AVA v2.2 (test)

39.8mAPUMT-L
23.1627.4831.836.12Dec 10, 2018Aug 28, 2019May 16, 2020Feb 1, 2021Oct 21, 2021Jul 9, 2022Mar 28, 2023
Updated 8mo ago

Evaluation results

MethodLinks
UMT-LPT Data=K710, Input Size=8×224², FLOPs (G)=596, Param (M)=3042023.0339.8
VideoMAE-LPT Data=K700, Input Size=16×224², FLOPs (G)=597, Param (M)=3052023.0339.3
UMT-LPT Data=K400, Input Size=8×224², FLOPs (G)=596, Param (M)=3042023.0339
MaskFeat-LPT Data=K600, Input Size=40×312², FLOPs (G)=2828, Param (M)=2182023.0337.8
ST-MAE-LPT Data=K700, Input Size=16×224², FLOPs (G)=598, Param (M)=3042023.0337.3
VideoMAE-LPT Data=K400, Input Size=16×224², FLOPs (G)=597, Param (M)=3052023.0337
MaskFeat-LPT Data=K400, Input Size=40×312², FLOPs (G)=2828, Param (M)=2182023.0336.3
ST-MAE-LPT Data=K400, Input Size=16×224², FLOPs (G)=598, Param (M)=3042023.0334.8
SlowFast++ ensembleBackbone=R101+NL, video pretrain=Kinetics-600, mode=ensemble2018.1234.3
SlowFastensemble=true, video pretrain=Kinetics-600, Backbone=R-101+NL, augmentation=multi-scale and horizontal flipping, region proposals=used for training2018.1234.3
MViTv2-LPT Data=IN-21K+K700, Input Size=40×312², FLOPs (G)=2828, Param (M)=2132023.0333.5
UMT-BPT Data=K710, Input Size=8×224², FLOPs (G)=180, Param (M)=872023.0333.5
UMT-BPT Data=K400, Input Size=8×224², FLOPs (G)=180, Param (M)=872023.0332.7
VideoMAE-BPT Data=K400, Input Size=16×224², FLOPs (G)=180, Param (M)=872023.0331.8
MViTv2-BPT Data=K700, Input Size=32×224², FLOPs (G)=225, Param (M)=512023.0331.3
MViTv1-BPT Data=K600, Input Size=32×224², FLOPs (G)=236, Param (M)=532023.0328.7
MViTv2-BPT Data=K400, Input Size=32×224², FLOPs (G)=225, Param (M)=512023.0328.1
SlowFastPT Data=K600, Input Size=64×224², FLOPs (G)=296, Param (M)=592023.0327.5
MViTv1-BPT Data=K400, Input Size=64×224², FLOPs (G)=455, Param (M)=362023.0327.3
SlowFastPT Data=K400, Input Size=32×224², FLOPs (G)=138, Param (M)=532023.0323.8