.png){kind=icon}
.png){kind=icon}
Return
Multi-Label Learning With Multiple Complementary Labels
.png){kind=icon}
PRE
AI
DOI：10.1109/TPAMI.2025.3574183.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In <i>multi-labeled complementary label learning</i> (MLCLL), a <i>complementary label</i> (CL) represents an irrelevant label for an instance. Utilizing CLs instead of relevant labels as annotations simplifies the annotation process in <i>multi-label learning</i> (MLL) tasks, underscoring the practicality of the MLCLL problem. However, existing MLCLL approaches mainly focus on scenarios where an instance is associated with a single CL. This restricts their applicability in situations where annotators provide multiple CLs per instance. To address this limitation, we propose a novel paradigm called multi-label learning with multiple complementary labels (ML-MCL), which allows each instance to be associated with multiple CLs simultaneously. Through analyzing the generation process of multiple CLs, we construct the relationship between relevant labels and CLs. This assists in deriving a tailored risk-consistent estimator to solve MLCLL with multiple CLs. Theoretically, we establish an estimation error bound for this estimator, with a convergence rate of <inline-formula><tex-math notation="LaTeX">$\mathcal {O}(1/\sqrt{n})$</tex-math></inline-formula>. Furthermore, we observed that unbounded gradients can be produced in the derived estimator when optimizing with certain loss functions, which may lead to unstable optimization. To mitigate this issue, we enhance the estimator with a <i>confidence truncation</i> loss, stabilizing the optimization process. Experimental results confirm the effectiveness of our approach, showing improved learning stability and performance in MLCLL tasks involving multiple CLs.
Keywords:
Complementary label learning
multi-label learning
risk-consistent estimator
estimation error bound
Journal
IF:
18.6
Papers:
831
Citations:
9.8W