Self Supervised Action Recognition On Ucf101

Metrics

3-fold Accuracy

Frozen

Pre-Training Dataset

Results

Performance results of various models on this benchmark

Model Name	3-fold Accuracy	Frozen	Pre-Training Dataset	Paper Title	Repository
3D RotNet (3D ResNet-18)	62.9	false	Kinetics400	Self-Supervised Spatiotemporal Feature Learning via Video Rotation Prediction	-
DPC (3D ResNet-18, Split 1)	60.6	false	UCF101	Video Representation Learning by Dense Predictive Coding
XKD-Modality-Agnostic (ViT-B/112/16)	93.4	-	-	XKD: Cross-modal Knowledge Distillation with Domain Alignment for Video Representation Learning
ViCC (S3D; R+F)	90.5	false	UCF101	Self-supervised Video Representation Learning with Cross-Stream Prototypical Contrasting
CVRL (R3D-50; K400)	92.2	false	Kinetics400	Spatiotemporal Contrastive Video Representation Learning
3D Cubic Puzzles (3D ResNet-18)	65.8	false	Kinetics400	Self-Supervised Video Representation Learning with Space-Time Cubic Puzzles	-
AVID (Modified R2+1D-18 on Audioset)	91.0	false	Audioset (Audio+Video)	Audio-Visual Instance Discrimination with Cross-Modal Agreement
CVRL (R3D-50; K600)	93.4	false	Kinetics600	Spatiotemporal Contrastive Video Representation Learning
CVRL (R3D-152 2x; K600)	93.9	false	Kinetics600	Spatiotemporal Contrastive Video Representation Learning
VideoGan (C3D)	52.1	false	UCF101	Generating Videos with Scene Dynamics	-
AVID+CMA (Modified R2+1D-18 on Kinetics)	87.5	false	Kinetics400 (Audio+Video)	Audio-Visual Instance Discrimination with Cross-Modal Agreement
RSPNet	93.7	false	Kinetics400	RSPNet: Relative Speed Perception for Unsupervised Video Representation Learning
ViCC (S3D; RGB)	72.2	true	UCF101	Self-supervised Video Representation Learning with Cross-Stream Prototypical Contrasting
VideoMAE(no extra data)	91.3	false	no extra data	VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training	-
AVID+CMA (Modified R2+1D-18 on Audioset)	91.5	false	Audioset (Audio+Video)	Audio-Visual Instance Discrimination with Cross-Modal Agreement
MVD (ViT-B)	97.5	false	Kinetics400	Masked Video Distillation: Rethinking Masked Feature Modeling for Self-supervised Video Representation Learning
CrissCross (Kinetics400)	91.5	false	Kinetics400	Self-Supervised Audio-Visual Representation Learning with Relaxed Cross-Modal Synchronicity
CrissCross (Kinetics-Sound)	88.3	false	Kinetics-Sound	Self-Supervised Audio-Visual Representation Learning with Relaxed Cross-Modal Synchronicity
ViCC (S3D; RGB)	88.8	false	UCF101	Self-supervised Video Representation Learning with Cross-Stream Prototypical Contrasting
ViCC (R2+1D; RGB)	82.8	false	UCF101	Self-supervised Video Representation Learning with Cross-Stream Prototypical Contrasting

0 of 53 row(s) selected.