Imbalance Learning-based Framework for Fear Recognition in the MediaEval Emotional Impact of Movies Task

Xiaotong Zhang, Xingliang Cheng, Mingxing Xu, Thomas Fang Zheng


Fear recognition, which aims at predicting whether a movie segment can induce fear or not, is a promising area in movie emotion recognition. Research in this area, however, has reached a bottleneck. Difficulties may partly result from the imbalanced database. In this paper, we propose an imbalance learning-based framework for movie fear recognition. A data rebalance module is adopted before classification. Several sampling methods, including the proposed softsampling and hardsampling which combine the merits of both undersampling and oversampling, are explored in this module. Experiments are conducted on the MediaEval 2017 Emotional Impact of Movies Task. Compared with the current state-of-the-art, we achieve an improvement of 8.94% on F1, proving the effectiveness of proposed framework.


 DOI: 10.21437/Interspeech.2018-1744

Cite as: Zhang, X., Cheng, X., Xu, M., Zheng, T.F. (2018) Imbalance Learning-based Framework for Fear Recognition in the MediaEval Emotional Impact of Movies Task. Proc. Interspeech 2018, 3678-3682, DOI: 10.21437/Interspeech.2018-1744.


@inproceedings{Zhang2018,
  author={Xiaotong Zhang and Xingliang Cheng and Mingxing Xu and Thomas Fang Zheng},
  title={Imbalance Learning-based Framework for Fear Recognition in the MediaEval Emotional Impact of Movies Task},
  year=2018,
  booktitle={Proc. Interspeech 2018},
  pages={3678--3682},
  doi={10.21437/Interspeech.2018-1744},
  url={http://dx.doi.org/10.21437/Interspeech.2018-1744}
}