Class-attribute inconsistency learning for novelty detection

作者：

Highlights：

• A new notion of class-attribute inconsistency for novelty detection.

• A novelty often has inconsistent class- and attribute-level similar references.

• CAILNet outperforms state-of-the-arts by exploring and mining inconsistency.

摘要

•A new notion of class-attribute inconsistency for novelty detection.•A novelty often has inconsistent class- and attribute-level similar references.•CAILNet outperforms state-of-the-arts by exploring and mining inconsistency.

论文关键词：Novelty detection,Class-attribute inconsistency,Class-level similarity,Attribute-level similarity

论文评审过程：Received 2 July 2021, Revised 25 November 2021, Accepted 8 February 2022, Available online 10 February 2022, Version of Record 17 February 2022.

论文官网地址：https://doi.org/10.1016/j.patcog.2022.108582