OpenSlot: Mixed Open-Set Recognition With Object-Centric Learning

doi:10.1109/TMM.2025.3565972

	OpenSlot: Mixed Open-Set Recognition With Object-Centric Learning
	Yin, Xu 1; Pan, Fei 2; An, Guoyuan 1; Huo, Yuchi 3,4; Xie, Zixuan 5; Yoon, Sung-Eui 6
	2025
发表期刊	IEEE TRANSACTIONS ON MULTIMEDIA
ISSN	1520-9210
卷号	27 页码:6019-6030
摘要	Existing open-set recognition (OSR) studies typically assume that each image contains only one class label,with the unknown test set (negative) having a disjoint label space from the known test set (positive), a scenario referred to as full-label shift. This paper introduces the mixed OSR problem, where test images contain multiple class semantics, with both known and unknown classes co-occurring in the negatives, leading to a more complex super-label shift that better reflects real-world scenarios. To tackle this challenge, we propose the OpenSlot framework, based on object-centric learning, which uses slot features to represent diverse class semantics and generate class predictions. The proposed anti-noise slot (ANS) technique helps mitigate the impact of noise (invalid or background) slots during classification training, addressing the semantic misalignment between class predictions and ground truth. We evaluate OpenSlot on both mixed and conventional OSR benchmarks. Without elaborate designs, our method not only excels existing approaches in detecting super-label shifts across OSR tasks, but also achieves state-of-the-art performance on conventional benchmarks. Meanwhile, OpenSlot can localize class objects without using bounding boxes during training, demonstrating competitive performance in open-set object detection and potential for generalization.
关键词	Semantics Training Object detection Benchmark testing Image reconstruction Dogs Safety Prototypes Object recognition Linear programming Mixed open-set recognition object-centric learning open-set object detection
DOI	10.1109/TMM.2025.3565972
收录类别	SCI
语种	英语
资助项目	Institute of Information & Communications Technology Planning & Evaluation (IITP) grant - Korea government (MSIT)[RS-2023-00237965] ; Recognition, Action and Interaction Algorithms for Open-world Robot Service ; National Research Foundation of Korea - Korea government (MSIT)[RS-2023-00208506] ; National Key Research and Development Program of China[2024YDLN0011] ; National Natural Science Foundation of China[62441205]
WOS研究方向	Computer Science ; Telecommunications
WOS类目	Computer Science, Information Systems ; Computer Science, Software Engineering ; Telecommunications
WOS记录号	WOS:001579069300032
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/41683
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Yoon, Sung-Eui
作者单位	1.Korea Adv Inst Sci & Technol, Sch Comp, Daejeon 34141, South Korea 2.Univ Michigan, Sch Comp Sci & Engn, Ann Arbor, MI 48109 USA 3.Zhejiang Univ, State Key Lab CAD & CG, Hangzhou 310027, Zhejiang, Peoples R China 4.Zhejiang Lab, Hangzhou 310058, Peoples R China 5.Univ Chinese Acad Sci, Inst Comp Technol, Beijing 100190, Peoples R China 6.Korea Adv Inst Sci & Technol, Fac Sch Comp, Deajeon 34141, South Korea
推荐引用方式 GB/T 7714	Yin, Xu,Pan, Fei,An, Guoyuan,et al. OpenSlot: Mixed Open-Set Recognition With Object-Centric Learning[J]. IEEE TRANSACTIONS ON MULTIMEDIA,2025,27:6019-6030.
APA	Yin, Xu,Pan, Fei,An, Guoyuan,Huo, Yuchi,Xie, Zixuan,&Yoon, Sung-Eui.(2025).OpenSlot: Mixed Open-Set Recognition With Object-Centric Learning.IEEE TRANSACTIONS ON MULTIMEDIA,27,6019-6030.
MLA	Yin, Xu,et al."OpenSlot: Mixed Open-Set Recognition With Object-Centric Learning".IEEE TRANSACTIONS ON MULTIMEDIA 27(2025):6019-6030.