This tool is used to label the dataset OA18. It can be used in general for any video action labeling (both action recognition/localization) given a long video. It is part of the study published in https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9014841
The videos are split into 2 parts: front view and side view perspectives.
- Part 1 - Front View Perspective: www.kaggle.com/basselsafwat/oa18-front-view
- Part 2 - Side View Perspective: www.kaggle.com/basselsafwat/oa18-side-view
The dataset is recorded with the cooperation of the theatre team of the faculty of Computer Science at Ain Shams University. The ultimate aim is to provide more data for the study of actions recognition and localization problem for the indoor office problem. I would like to thank them all for their time and performance which looked very realistic while recording this dataset.
If you use this tool or the dataset or part of it in your work, you would need to cite the original work as follow:
Chawkv BS, Marey M, Shedeed HA. OA18: A New Office Actions Benchmark. In 2019 Ninth International Conference on Intelligent Computing and Information Systems (ICICIS) 2019 Dec 8 (pp. 32-38). IEEE.
Feel free to contact me on [email protected]