List your company
ResearchFormerly WizWand

Action Recognition on UCF101 (Split 1)

96.8Top-1 AccSupervised
70.59277.39684.291.004Nov 28, 2019Dec 31, 2020Feb 3, 2022Mar 9, 2023Apr 11, 2024May 15, 2025Jun 19, 2026
Updated 4mo ago

Evaluation results

MethodLinks
SupervisedBackbone=R(2+1)D-18 (31.5), Pre-trained Dataset=AudioSet, Finetune input=32 x 224^22021.1196.8——
Supervised PretrainingPretrain=Kinetics, Backbone=R(2+1)D, Strategy=Supervised Pretraining, N x H/W=16x1122026.0696.8——
VicTRk (shots)=16, Pre-training=w/ image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=50, Backbone=B/162023.0495.8——
GDTEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=IG65M (21y), Modality=VA2020.0895.2——
MMVEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=AS+HT (16y), Modality=VAT2020.0895.2——
SupervisedBackbone=R(2+1)D-18 (31.5), Pre-trained Dataset=Kinetics400, Finetune input=32 x 224^22021.1195——
X-Florencek (shots)=16, Pre-training=w/ image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=502023.0494.8——
XDCEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=IG65M (21y), Modality=VA2020.0894.2——
CVRLEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K600 (44d), Modality=V, Backbone=R3D-152(2x)2020.0893.9——
EloEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=YT8M (13y), Modality=VFA2020.0893.8——
BraVeCompute=16 TPUs, Backbone=R(2+1)D-18 (31.5), Pre-trained Dataset=AudioSet, Finetune input=32 x 224^2, Note=pretrained with very high temporal resolutions (2 views of 32 & 128 frames)2021.1193.6——
VicTRk (shots)=8, Pre-training=w/ image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=50, Backbone=B/162023.0493.6——
CM-ACCCompute=40 GPUs, Backbone=R(2+1)D-18 (33.4), Pre-trained Dataset=AudioSet, Finetune input=32 x 224^22021.1193.5——
CVRLEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K600 (44d), Modality=V, Backbone=R3D-502020.0893.4——
XDCCompute=64 GPUs, Backbone=R(2+1)D-18 (31.5), Pre-trained Dataset=AudioSet, Finetune input=32 x 224^22021.1193——
GDTEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=AS (240d), Modality=VA2020.0892.5——
X-Florencek (shots)=8, Pre-training=w/ image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=502023.0492.5——
CrissCrossCompute=8 GPUs, Backbone=R(2+1)D-18 (15.4), Pre-trained Dataset=AudioSet, Finetune input=32 x 224^22021.1192.4——
VicTRk (shots)=4, Pre-training=w/ image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=50, Backbone=B/162023.0492.3——
CVRLEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=V, Backbone=R3D-502020.0892.2——
MMVEvaluation Protocol=Linear Evaluation, Pre-train data (duration)=AS+HT (16y), Modality=VAT2020.0891.8——
XDCPretraining Dataset=IG-Kinetics, Pretraining Size=65M2019.1191.5——
CrissCrossCompute=8 GPUs, Backbone=R(2+1)D-18 (15.4), Pre-trained Dataset=Kinetics400, Finetune input=32 x 224^22021.1191.5——
MMVCompute=32 TPUs, Backbone=R(2+1)D-18 (31.5), Pre-trained Dataset=AudioSet, Finetune input=32 x 224^22021.1191.5——
AVIDCompute=64 GPUs, Backbone=R(2+1)D-18 (15.4), Pre-trained Dataset=AudioSet, Finetune input=32 x 224^22021.1191.5——
X-CLIPk (shots)=16, Pre-training=w/ image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=502023.0491.4——
MIL-NCEEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=HT (15y), Modality=VT2020.0891.3——
XDCEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=AS (240d), Modality=VA2020.0891.2——
MMV-VAEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=AS+HT (16y), Modality=VA2020.0891.1——
SupervPretraining Dataset=Kinetics, Pretraining Size=240K2019.1190.9——
GDTCompute=64 GPUs, Backbone=R(2+1)D-18 (31.5), Pre-trained Dataset=Kinetics400, Finetune input=32 x 224^22021.1190.9——
CoCLREvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=VF2020.0890.6——
CVRLEvaluation Protocol=Linear Evaluation, Pre-train data (duration)=K600 (44d), Modality=V, Backbone=R3D-502020.0890.6——
COCLR†Input Size=32 x 128x128, Arch=S3D-23, Modality=RGB + Optical Flow2022.0690.6——
CMACCompute=8 GPUs, Backbone=R(2+1)D-18 (31.5), Pre-trained Dataset=Kinetics400, Finetune input=32 x 224^22021.1190.3——
CM-ACCCompute=40 GPUs, Backbone=3D-ResNet18 (33.4), Pre-trained Dataset=Kinetics400, Finetune input=32 x 224^2, Dataset samples=240K samples from Kinetics7002021.1190.2——
CrissCrossCompute=8 GPUs, Backbone=R(2+1)D-18 (15.4), Pre-trained Dataset=AudioSet, Finetune input=8 x 224^22021.1189.4——
TimeSformerk (shots)=16, Pre-training=w/o image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=502023.0489.4——
CVRLEvaluation Protocol=Linear Evaluation, Pre-train data (duration)=K400 (28d), Modality=V, Backbone=R3D-502020.0889.2——
AVTSEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=AS (240d), Modality=VA2020.0889——
XDCPretraining Dataset=IG-Random, Pretraining Size=65M2019.1188.8——
Video-Swin-Bk (shots)=16, Pre-training=w/o image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=50, Backbone=Swin-B2023.0488.7——
AVIDCompute=64 GPUs, Backbone=R(2+1)D-18 (15.4), Pre-trained Dataset=AudioSet, Finetune input=8 x 224^22021.1188.6——
X-Florencek (shots)=4, Pre-training=w/ image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=502023.0488.5——
CrissCrossCompute=4 GPUs, Backbone=R(2+1)D-18 (15.4), Pre-trained Dataset=Kinetics-Sound, Finetune input=32 x 224^22021.1188.3——
X-CLIPk (shots)=8, Pre-training=w/ image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=502023.0488.3——
DynamoNetEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=YT8M (13y), Modality=V2020.0888.1——
COCLR-RGBInput Size=32 x 128x128, Arch=S3D-23, Modality=RGB2022.0687.9——
Supervised PretrainingPretrain=Kinetics, Backbone=R3D-18, Strategy=Supervised Pretraining, N x H/W=16x1122026.0687.8——
VicTRk (shots)=2, Pre-training=w/ image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=50, Backbone=B/162023.0487.7——
AVIDCompute=64 GPUs, Backbone=R(2+1)D-18 (15.4), Pre-trained Dataset=Kinetics400, Finetune input=32 x 224^22021.1187.5——
SupervisedBackbone=3D-ResNet18 (33.4), Pre-trained Dataset=Kinetics-Sound, Finetune input=32 x 224^22021.1186.9——
CrissCrossCompute=8 GPUs, Backbone=R(2+1)D-18 (15.4), Pre-trained Dataset=Kinetics400, Finetune input=8 x 224^22021.1186.9——
XDCCompute=64 GPUs, Backbone=R(2+1)D-18 (31.5), Pre-trained Dataset=Kinetics400, Finetune input=32 x 224^22021.1186.8——
SLIC†Input Size=32 x 128x128, Arch=R3D-18, Modality=RGB + Optical Flow, Pretraining Frames=162022.0686.3——
MMV-VAEvaluation Protocol=Linear Evaluation, Pre-train data (duration)=AS+HT (16y), Modality=VA2020.0886.2——
Video-Swin-Bk (shots)=8, Pre-training=w/o image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=50, Backbone=Swin-B2023.0485.8——
DistInitPretrain=Kinetics+Sports-1M, Backbone=R(2+1)D-18, Strategy=DistInit [6], N x H/W=32x1122026.0685.7——
Robust-xIDCompute=8 GPUs, Backbone=R(2+1)D-18 (15.4), Pre-trained Dataset=Kinetics400, Finetune input=32 x 224^22021.1185.6——
XDCEvaluation Protocol=Linear Evaluation, Pre-train data (duration)=AS (240d), Modality=VA2020.0885.3——
XDCPretraining Dataset=AudioSet, Pretraining Size=2M2019.1184.9——
XDCCompute=64 GPUs, Backbone=R(2+1)D-18 (31.5), Pre-trained Dataset=AudioSet, Finetune input=8 x 224^22021.1184.9——
TCLRPretrain=Kinetics, Backbone=R3D-18, Strategy=TCLR [27], N x H/W=16x1122026.0684.1——
SupervPretraining Dataset=AudioSet, Pretraining Size=2M2019.1184——
X-Florencek (shots)=2, Pre-training=w/ image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=502023.0484——
AVIDCompute=64 GPUs, Backbone=R(2+1)D-18 (15.4), Pre-trained Dataset=Kinetics400, Finetune input=8 x 224^22021.1183.7——
TimeSformerk (shots)=8, Pre-training=w/o image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=502023.0483.7——
MIL-NCEEvaluation Protocol=Linear Evaluation, Pre-train data (duration)=HT (15y), Modality=VT2020.0883.4——
X-CLIPk (shots)=4, Pre-training=w/ image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=502023.0483.4——
SLICInput Size=32 x 128x128, Arch=R3D-18, Modality=RGB, Pretraining Frames=162022.0683.2——
SeLaViCompute=64 GPUs, Backbone=R(2+1)D-18 (31.5), Pre-trained Dataset=Kinetics400, Finetune input=32 x 224^22021.1183.1——
TCLRPretrain=UCF, Backbone=R3D-18, Strategy=TCLR [27], N x H/W=16x1122026.0682.4——
Supervised PretrainingPretrain=Sports-1M, Backbone=C3D, Strategy=Supervised Pretraining, N x H/W=16x1122026.0682.3——
Robust-xIDCompute=8 GPUs, Backbone=R(2+1)D-18 (15.4), Pre-trained Dataset=Kinetics400, Finetune input=8 x 224^22021.1181.9——
Supervised PretrainingPretrain=Kinetics, Backbone=R3D-18, Strategy=Supervised Pretraining, N x H/W=8x2242026.0681.7——
SpeedNetEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=V2020.0881.1——
SpeedNetPretrain=Kinetics, Backbone=S3D-G, Strategy=SpeedNet [22], N x H/W=64x2242026.0681.1——
SupervPretraining Dataset=ImageNet, Pretraining Size=1.2M2019.1179.9——
CBTEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K600+ (273d), Modality=V2020.0879.5——
VideoMoCoPretrain=Kinetics, Backbone=R(2+1)D-18, Strategy=VideoMoCo [26], N x H/W=32x1122026.0678.7——
CoCon-E*†Input Size=224x224, Arch=R3D-18, Modality=RGB + Optical Flow2022.0678.1——
CoCLREvaluation Protocol=Linear Evaluation, Pre-train data (duration)=K400 (28d), Modality=VF2020.0877.8——
XDCPretraining Dataset=AudioSet-240K, Pretraining Size=240K2019.1177.4——
CM-ACCCompute=40 GPUs, Backbone=3D-ResNet18 (33.4), Pre-trained Dataset=Kinetics-Sound, Finetune input=32 x 224^22021.1177.2——
PacePredEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=V2020.0877.1——
PacePredPretrain=Kinetics, Backbone=R(2+1)D-18, Strategy=PacePred [44], N x H/W=16x1122026.0677.1——
DANetPretrain=UCF/HMDB, Backbone=R3D-18, Strategy=DANet [5], N x H/W=8x1122026.0676.8——
SupervPretraining Dataset=AudioSet-240K, Pretraining Size=240K2019.1176.6——
X-CLIPk (shots)=2, Pre-training=w/ image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=502023.0476.4——
DPCEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=V2020.0875.7——
TimeSformerk (shots)=4, Pre-training=w/o image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=502023.0475.6——
ImageNet InflationPretrain=None, Backbone=R3D-18, Strategy=ImageNet Inflation, N x H/W=16x1122026.0674.3——
XDCPretraining Dataset=Kinetics, Pretraining Size=240K2019.1174.2——
XDCCompute=64 GPUs, Backbone=R(2+1)D-18 (31.5), Pre-trained Dataset=Kinetics400, Finetune input=8 x 224^22021.1174.2——
Video-Swin-Bk (shots)=4, Pre-training=w/o image-text, Pre-training dataset=Kinetics-400, Finetuning epochs=50, Backbone=Swin-B2023.0474.1——
VideoMoCoPretrain=Kinetics, Backbone=R3D-18, Strategy=VideoMoCo [26], N x H/W=32x1122026.0674.1——
LEViLPretrain=UCF+HMDB+Kineticsb, Backbone=R3D-18, Strategy=LEViLc, N x H/W=8x2242026.0672.5——
ClipOrderEvaluation Protocol=Fine-Tuning, Pre-train data (duration)=K400 (28d), Modality=V2020.0872.4——
ClipOrderPretrain=UCF, Backbone=R(2+1)D-18, Strategy=ClipOrder [25], N x H/W=16x1122026.0672.4——
CoCon-RGB*Input Size=224x224, Arch=R3D-18, Modality=RGB2022.0671.6——