This tool, written in Python as a Flask application backed with a MongoDB, allows any number people to annotate video clips per-frame, for valence and arousal, remotely. In addition it can be easily extended to handle more annotations (e.g. discrete emotions).
The tool is open source and can be obtained on github
, along with instructions.
If you use this annotation tool or a modified version, please cite:
* J. Kossaifi, G. Tzimiropoulos, S. Todorovic and M. Pantic. AFEW-VA for valence and arousal estimation In-The-Wild. Image and Vision Computing, 2016 (submitted).