anomaly detection can be conceived either through generative modelling of
regular training data or by discriminating with respect to negative training
data. These two approaches exhibit different failure modes. Consequently,
hybrid algorithms present an attractive research goal. Unfort
本研究提出了 OpenHybrid 框架,应用于 open set recognition 中,在 inlier 分类器和 density estimator 之间进行联合学习,通过 encoder 将输入数据编码为联合嵌入空间,使用 flow-based 模型检测未知类别的样本是否属于异常值,并且实验结果表明,该模型在标准 open set 基准上显著优于现有的方法和基线模型。