zhangchengpu/CMOSE

CMOSE: Comprehensive Multi-Modality Online Student Engagement Dataset with High-Quality Labels Project page is here Video clip name Each video clip is named as videoX_Y_personZ, which means it is the Yth clip of the Zth subject from coaching session X. Openface We extract the second level features from OpenFace. The extracted files are stored under "secondfeature/videoX_Y_personZ.csv". These features include: Gaze Direction and… See the full description on the dataset page: https://huggingface.co/datasets/zhangchengpu/CMOSE.