This repository contains Youtube Gesture-Type Dataset and scripts for dividing videos.
You can download TED Gesture-Type Dataset from here
You can load the dataset with the following code.
import pickle
data = pickle.load(open('TED_gesture-type_dataset.pickle', "rb"))The data is dictionary type, with 9 variables, each containing 13,736 gestures.
- vid_id : YouTube video ID
- gesture_id : Gesture ID
- gesture_type : Gesture types. This is annotated with either Beat, Imagistic or No-Gesture.
- start_frame_no : Gesture start frame in YouTube video
- end_frame_no : Gesture end frame in YouTube video
- text : speech sentence
- remark : Words expressed as gestures in text. This is annotated only for Imagistic.
- word_timestamp : Word and the time it was uttered.
- skeletons : 3D pose sequence. The definition of joints follows the Kinect format as shown in the image below.
If you need videos for each gesture, run the following script. Install ffmpeg and verify that the data path is correct in the script.
pip install -r requirements.txt
python youtube_dl.py
python segment_videos.py
@misc{https://doi.org/10.48550/arxiv.2210.06790,
doi = {10.48550/ARXIV.2210.06790},
url = {https://arxiv.org/abs/2210.06790},
author = {Teshima, Hitoshi and Wake, Naoki and Thomas, Diego and Nakashima, Yuta and Kawasaki, Hiroshi and Ikeuchi, Katsushi},
title = {Deep Gesture Generation for Social Robots Using Type-Specific Libraries},
publisher = {arXiv},
year = {2022}
}
This dataset is based on the YouTube Gesture Dataset by Yoon et al.
@INPROCEEDINGS{
yoonICRA19,
title={Robots Learn Social Skills: End-to-End Learning of Co-Speech Gesture Generation for Humanoid Robots},
author={Yoon, Youngwoo and Ko, Woo-Ri and Jang, Minsu and Lee, Jaeyeon and Kim, Jaehong and Lee, Geehyuk},
booktitle={Proc. of The International Conference in Robotics and Automation (ICRA)},
year={2019}
}