Action Recognition In Videos

Best model: Text4Vis (w/ ViT-L)

Best model: VideoMAE V2-g

Best model: MSQNet

Best model: PoseC3D (RGB + Pose)

Best model: PoseC3D (RGB + Pose)

Best model: MVD (Kinetics400 pretrain, ViT-H, 16 frame)

Best model: InternVideo

Best model: ip-CSN-152 (RGB)

Best model: BMN

Best model: R3D-18

Best model: VideoMAE V2-g

Best model: PoseC3D (Pose Only)

Best model: LART (Hiera-H, K700 PT+FT)

Best model: LaViLa (Finetuned, TimeSformer-L)

Best model: Avion (ViT-L)

Best model: HandFormer-B/21x8

Best model: UniFormerV2-L

Best model: DirecFormer

Best model: CNN-GRU

Best model: SlowFast

Best model: JMRN

Best model: C3D-AVG

Best model: DVANet

Best model: DeVTr

Best model: SEW-Resnet18 (3sets)

Best model: PMI Sampler

Best model: FAR

Best model: R2+1D-BERT

Best model: 2DCNN+TRN