Emotionally Evocative Video DB

EVAT: database of Emotionally evocative Video clips and their associated Affective Tagging

Daniel Hadar

Hebrew University, Jerusalem

11 January 2017

This page includes download links to EVAT, a tagged database of 36 short emotionally evocative video clips.
The database is composed of 2 files:

videos_list.xlsx: A list of the 36 video clips including name, YouTube link, full duration, and the segment used for tagging (defined by start and end time).
all_clips.csv: The affective ratings of 26 participants when viewing the video clips, using the following scales: Valence, Arousal, Likability, Rewatch (the desire to watch again) and Familiarity.

If you decide to make use of EVAT in your work, please kindly add a citation to our paper (see below).
To download use this link.

Distribution of 4 subjective scores over all tested clips, where Valence and Arousal define the two main axes, also summarized in histogram form above and to the right of the plot. Size and color correspond respectively to the remaining 2 scores of Rewatch and Likability.
All video clips were found significantly unfamiliar among the 26 participants (p < 0.0001).

Citation: Daniel Hadar, Talia Tron and Daphna Weinshall. Implicit Media Tagging and Affect Prediction from RGB-D video of spontaneous facial expressions. Proceedings: 12th IEEE International Conference on Automatic Face and Gesture Recognition (FG2017), Washington DC, May 2017.
For further details, feel free to contact me at: daniel.hadar at mail.huji.ac.il