Papperoni

2012

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

Khurram Soomro, Amir Roshan Zamir, Mubarak Shah

citations

Cite Score

AI summary

This paper introduces UCF101, a new large-scale dataset for human action recognition from web videos, featuring 101 action classes, over 13k clips, and 27 hours of data, and provides baseline recognition results using a bag-of-words approach.

Main Contributions

Introduction of UCF101, the largest dataset for human action recognition, with 101 action classes, over 13,000 video clips, and 27 hours of video data.
UCF101 features realistic user-uploaded videos, presenting challenges like camera motion and cluttered backgrounds, making it a highly unconstrained dataset.
Provides baseline action recognition results on UCF101 using a standard bag-of-words approach, achieving an overall performance of 44.5%.
Categorization of action classes into five types: Human-Object Interaction, Body-Motion Only, Human-Human Interaction, Playing Musical Instruments, and Sports.
Highlights that sports actions achieve the highest accuracy due to distinctive motions and less cluttered backgrounds.

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.

Citation Graph

Loading graph...

References [12]

Sort:

Filter:

[1]Actions as Space-Time Shapes

M. Blank, L. Gorelick, E. Shechtman, M. Irani, R. Basri - 2005

2 papers in library cite

Google Scholar

[2]Modeling Temporal Structure of Decomposable Motion Segments for Activity Classification

J. C. Niebles, C. W. Chen, Li Fei Fei - 2010

2 papers in library cite