UCF101 (UCF101 Human Actions dataset)

Introduced by Soomro et al. in UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild

UCF101 dataset is an extension of UCF50 and consists of 13,320 video clips, which are classified into 101 categories. These 101 categories can be classified into 5 types (Body motion, Human-human interactions, Human-object interactions, Playing musical instruments and Sports). The total length of these video clips is over 27 hours. All the videos are collected from YouTube and have a fixed frame rate of 25 FPS with the resolution of 320 × 240.

Source: Two-stream Collaborative Learning with Spatial-Temporal Attention for Video Classification

Homepage

Benchmarks

Add a new result Link an existing benchmark

Task	Dataset Variant	Best Model
Action Recognition	UCF101	VideoMAE V2-g
Self-Supervised Action Recognition	UCF101	VideoMAE V2-g
Video Generation	UCF-101	W.A.L.T-XL
Zero-Shot Action Recognition	UCF101	OTI
Video Frame Interpolation	UCF101	MISO-VFI
Self-Supervised Action Recognition	UCF101 (finetuned)	CVRL
Self-supervised Video Retrieval	UCF101	SCE
Text-to-Video Generation	UCF-101	REGIS-Fuse
Prompt Engineering	UCF101	PromptKD
Video Generation	UCF-101 16 frames, 64x64, Unconditional	Make-A-Video
Few Shot Action Recognition	UCF101	STRM
Video Generation	UCF-101 16 frames, Unconditional, Single GPU	TGAN-F
Video Generation	UCF-101 16 frames, 128x128, Unconditional	TGANv2
Action Recognition In Videos	UCF101	STM
Action Recognition	UCF-101	DMC-Net
Few-Shot Learning	UCF101	Variational Prompt Tuning
Open Set Action Recognition	UCF101-MiTv2	InternVideo
Skeleton Based Action Recognition	UCF101	Structured Keypoint Pooling
Action Classification	UCF101	Ours
Action Recognition	UCF 101	R2+1D-BERT
Self-Supervised Action Recognition Linear	UCF101	XKD
Early Action Prediction	UCF101	TemPr4