AgoraComa Behaviour Analysis

This dataset contains behavioral data of a control group of 16 persons undergoing different emotional and neutral stimulations (sound, reading of an auto-biographical text etc.). It is the dataset used for the experiments in [2]. Face and body landmarks have been extracted from videos using OpenPose and OpenFace and post-processed and normalised according to the procedure explained in [1]. Reference : [1] Michelot, B., Corneyllie, A., Thevenet, M., Duffner S., Perrin, F. A modular machine learning tool for holistic and fine-grained behavioral analysis. Behav Res 57, 24 (2025). https://doi.org/10.3758/s13428-024-02511-3 [2] Michelot, B., Perrin, F., Duffner S. Explaining time series classifications for understanding sparse human behavioural patterns in naturalistic situations