ActionDataset
Description
Action Dataset Prepared by: Haoran Xu (haoran.x@gmail.com) Desimone Lab, McGovern Institute for Brain Research, MIT If you use this dataset, please cite: Rajimehr R., Xu H., Farahani A., Kornblith S., Duncan J., Desimone R. Functional architecture of cerebral cortex during naturalistic movie-watching. bioRxiv: https://www.biorxiv.org/content/10.1101/2022.03.14.483878v1 doi: https://doi.org/10.1101/2022.03.14.483878 The dataset includes 5 action categories: H: human action HH: human-human interaction HO: human-object interaction O: object motion OO: object-object interaction Each category contains 10 subcategories, and each subcategory contains 6 example stimuli (300 stimuli in total). The duration of all video clips is 4.96 seconds (124 frames at a frame rate of 25 fps). Some of the action videos were originally from the UCF101 dataset (https://www.crcv.ucf.edu/data/UCF101.php), and the remaining videos were obtained from YouTube.