OMG-Emotion stands for One-Minute Gradual-Emotional Behavior dataset. The dataset contains 567 emotional videos with an average length of 1 minute, and 7,371 video clips segmented and annotated based on discourse (each discourse has at least 5 independent topics). The videos were selected using crawler technology, searching based on long-term emotional behaviors such as “monologue,” “audition,” “dialogue,” and “emotional scene.”

Citation

@inproceedings{Barros2018OMG, author={Barros, Pablo and Churamani, Nikhil and Lakomkin, Egor and Sequeira, Henrique and Sutherland, Alexander and Wermter, Stefan}, booktitle={2018 International Joint Conference on Neural Networks (IJCNN)}, title={{The OMG-Emotion Behavior Dataset}}, year={2018}, moth={July}, organization={IEEE}, pages={1408–1414}, location={Rio de Janeiro, Brazil}, doi={10.1109/IJCNN.2018.8489099} }