UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild

Abstract

We introduce UCF101 which is currently the largest dataset of human actions. It consists of 101 action classes, over 13k clips and 27 hours of video data. The database consists of realistic user uploaded videos containing camera motion and cluttered background. Additionally, we provide baseline action recognition results on this new dataset using standard bag of words approach with overall performance of 44.5%. To the best of our knowledge, UCF101 is currently the most challenging dataset of actions due to its large number of classes, large number of clips and also unconstrained nature of such clips.

Cited by in corpus (17)

• Towards Good Practices for Very Deep Two-Stream ConvNets
• Beyond Short Snippets: Deep Networks for Video Classification
• Evaluating Two-Stream CNN for Video Classification
• Exploiting Image-trained CNN Architectures for Unconstrained Video Classification
• Finding Action Tubes
• Temporal Pyramid Pooling Based Convolutional Neural Networks for Action Recognition
• Initialization Strategies of Spatio-Temporal Convolutional Neural Networks
• Modeling Spatial-Temporal Clues in a Hybrid Deep Learning Framework for Video Classification
• Watch and Learn: Semi-Supervised Learning of Object Detectors from Videos
• Beyond Gaussian Pyramid: Multi-skip Feature Stacking for Action Recognition
• Fine-grained Activity Recognition with Holistic and Pose based Features
• The Best of Both Worlds: Combining Data-independent and Data-driven Approaches for Action Recognition
• Learning Temporal Embeddings for Complex Video Analysis
• Pooled Motion Features for First-Person Videos
• Compositional Structure Learning for Action Understanding
• Articulated motion discovery using pairs of trajectories
• Gradient Boundary Histograms for Action Recognition