Yu Jongmin, Lee Younkwan, Yow Kin Choong, Jeon Moongu, Pedrycz Witold
IEEE Trans Neural Netw Learn Syst. 2022 Aug;33(8):3572-3586. doi: 10.1109/TNNLS.2021.3053563. Epub 2022 Aug 3.
We present adversarial event prediction (AEP), a novel approach to detecting abnormal events through an event prediction setting. Given normal event samples, AEP derives the prediction model, which can discover the correlation between the present and future of events in the training step. In obtaining the prediction model, we propose adversarial learning for the past and future of events. The proposed adversarial learning enforces AEP to learn the representation for predicting future events and restricts the representation learning for the past of events. By exploiting the proposed adversarial learning, AEP can produce the discriminative model to detect an anomaly of events without complementary information, such as optical flow and explicit abnormal event samples in the training step. We demonstrate the efficiency of AEP for detecting anomalies of events using the UCSD-Ped, CUHK Avenue, Subway, and UCF-Crime data sets. Experiments include the performance analysis depending on hyperparameter settings and the comparison with existing state-of-the-art methods. The experimental results show that the proposed adversarial learning can assist in deriving a better model for normal events on AEP, and AEP trained by the proposed adversarial learning can surpass the existing state-of-the-art methods.
我们提出了对抗性事件预测（AEP），这是一种通过事件预测设置来检测异常事件的新方法。给定正常事件样本，AEP会推导预测模型，该模型可以在训练步骤中发现事件当前与未来之间的相关性。在获取预测模型时，我们针对事件的过去和未来提出了对抗学习。所提出的对抗学习促使AEP学习用于预测未来事件的表示，并限制对事件过去的表示学习。通过利用所提出的对抗学习，AEP可以生成判别模型，在没有补充信息（如光流和训练步骤中的显式异常事件样本）的情况下检测事件异常。我们使用UCSD-Ped、CUHK Avenue、Subway和UCF-Crime数据集证明了AEP在检测事件异常方面的效率。实验包括根据超参数设置进行性能分析以及与现有最先进方法的比较。实验结果表明，所提出的对抗学习有助于在AEP上为正常事件推导更好的模型，并且通过所提出的对抗学习训练的AEP可以超越现有最先进方法。
