List your company
ResearchFormerly WizWand

Action Classification on HMDB51 (split1)

75AccuracyMMV
13.22429.26245.361.338Nov 21, 2016Oct 27, 2017Oct 2, 2018Sep 8, 2019Aug 13, 2020Jul 19, 2021Jun 25, 2022
Updated 8mo ago

Evaluation results

MethodLinks
MMVEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=AS+HT (16y), Modality=VAT2020.0875
Kinetics pretrained I3DModel Type=Multi-frame, Pre-training=Kinetics, Fine-tuned from=UCF1012016.1274.8
GDTEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=IG65M (21y), Modality=VA2020.0872.8
CVRLEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K600 (44d), Modality=V, Backbone=R3D-152(2x)2020.0869.9
MMV-VAEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=AS+HT (16y), Modality=VA2020.0868.3
CVRLEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K600 (44d), Modality=V, Backbone=R3D-502020.0868
XDCEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=IG65M (21y), Modality=VA2020.0867.4
EloEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=YT8M (13y), Modality=VFA2020.0867.4
MMVEvaluation Protocol=Linear Evaluation, Pre-train data (duration)=AS+HT (16y), Modality=VAT2020.0867.1
CVRLEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=V, Backbone=R3D-502020.0866.7
GDTEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=AS (240d), Modality=VA2020.0866.1
EloEvaluation Protocol=Linear Evaluation, Pre-train data (duration)=YT8M (13y), Modality=VFA2020.0864.5
CoCLREvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=VF2020.0862.9
COCLR†Input Size=32 x 128x128, Arch=S3D-23, Modality=RGB + Optical Flow2022.0662.9
AVTSEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=AS (240d), Modality=VA2020.0861.6
MMV-VAEvaluation Protocol=Linear Evaluation, Pre-train data (duration)=AS+HT (16y), Modality=VA2020.0861.5
MIL-NCEEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=HT (15y), Modality=VT2020.0861
XDCEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=AS (240d), Modality=VA2020.0861
DynamoNetEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=YT8M (13y), Modality=V2020.0859.9
CVRLEvaluation Protocol=Linear Evaluation, Pre-train data (duration)=K600 (44d), Modality=V, Backbone=R3D-502020.0859.7
CVRLEvaluation Protocol=Linear Evaluation, Pre-train data (duration)=K400 (28d), Modality=V, Backbone=R3D-502020.0857.3
SLIC†Input Size=32 x 128x128, Arch=R3D-18, Modality=RGB + Optical Flow, Pretraining Frames=162022.0657.3
ActionFlowNet (UCF101)Model Type=Multi-frame, Pre-training=FlowNet part pretrained on UCF101, Fine-tuned from=UCF1012016.1256.4
XDCEvaluation Protocol=Linear Evaluation, Pre-train data (duration)=AS (240d), Modality=VA2020.0856
MIL-NCEEvaluation Protocol=Linear Evaluation, Pre-train data (duration)=HT (15y), Modality=VT2020.0854.8
COCLR-RGBInput Size=32 x 128x128, Arch=S3D-23, Modality=RGB2022.0654.6
TSN, BN-inception (RGB)Backbone=BN-inception, Stream=RGB2017.1154.4
Pose regularized Attentional PoolingStream=RGB, Input Resolution=480px2017.1154.4
Sports-1M pretrained C3DModel Type=Multi-frame, Pre-training=Sports-1M, Fine-tuned from=UCF1012016.1253.5
CoCLREvaluation Protocol=Linear Evaluation, Pre-train data (duration)=K400 (28d), Modality=VF2020.0852.4
SLICInput Size=32 x 128x128, Arch=R3D-18, Modality=RGB, Pretraining Frames=162022.0652.2
CoCon-E*†Input Size=224x224, Arch=R3D-18, Modality=RGB + Optical Flow2022.0652
ActionVLADStream=RGB2017.1151.2
Linear Attentional PoolingStream=RGB, Input Resolution=480px2017.1151.1
Multi-frame FlowNet fine-tuneModel Type=Multi-frame, Initialization=Pretrained on UCF101 for optical flow, Fine-tuned from=UCF1012016.1250.6
SpeedNetEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=V2020.0848.8
TSN, ResNet101 (RGB)Backbone=ResNet101, Stream=RGB, Input Resolution=480px2017.1148.2
ImageNet pretrained ResNet-18Model Type=Two-frame, Initialization=ImageNet, Fine-tuned from=UCF1012016.1247.1
CoCon-RGB*Input Size=224x224, Arch=R3D-18, Modality=RGB2022.0646
CBTEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K600+ (273d), Modality=V2020.0844.6
ActionFlowNet-2F (FlCh+UCF101)Model Type=Two-frame, Pre-training=FlowNet part pretrained on FlyingChairs and UCF101, Fine-tuned from=UCF1012016.1242.6
StackedModel Type=Two-frame, Initialization=FlowNet pretrained on UCF101, Fine-tuned from=UCF1012016.1242.4
ActionFlowNet-2F (UCF101)Model Type=Two-frame, Pre-training=FlowNet part pretrained on UCF101, Fine-tuned from=UCF1012016.1242.4
Stack-of-Diff.Weights=ImageNet supervised2016.1140.8
PacePredEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=V2020.0836.6
DPCEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=V2020.0835.7
3D-RotNetEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=V2020.0834.3
ST-PuzzleEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=V2020.0833.7
MotionPredEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=V2020.0833.4
O3NEncoder=Stack-of-Diff.2016.1132.5
ClipOrderEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=V2020.0830.9
FlowNet fine-tuneModel Type=Two-frame, Initialization=Pretrained on UCF101 for optical flow, Fine-tuned from=UCF1012016.1229.1
Stack-of-Diff.Weights=Random2016.1128.3
ScratchModel Type=Two-frame, Backbone=ResNet-18, Initialization=Random, Fine-tuned from=UCF1012016.1223.9
Seq. Ver.2016.1119.8
DrLim2016.1116.3
TempCoh2016.1115.9
Obj. Patch2016.1115.6