Incomplete label distribution learning via label correlation decomposition

doi:10.1016/j.inffus.2024.102600

Incomplete label distribution learning via label correlation decomposition

2025-01-01

0

PRE

AI

DOI：10.1016/j.inffus.2024.102600

Original

Original request for help

Save

Abstract

En 中文

Label distribution learning (LDL) has garnered increased attention in recent studies on label ambiguity. However, collecting complete annotations for LDL tasks is often time-consuming and labor-intensive compared to traditional learning paradigms. Therefore, designing effective incomplete LDL algorithms is crucial to broaden LDL's application scope. In this paper, we propose a novel LDL algorithm, called Incom plete L abel D istribution L earning via L abel C orrelation D ecomposition (IncomLDL-LCD), which simultaneously learns label distributions and recovers missing description degrees of labels through label correlations. Specifically, we decompose the label correlation into sparse local label correlation and low-rank global label correlation using a softthresholding operator and a singular value thresholding operator, respectively. The former is utilized to capture the related label subsets necessary for reconstructing each possible label, while the latter focuses on extracting the coarse-grained semantic concepts from all labels and exploring the groupings of labels. Additionally, we develop an alternating solution with the accelerated proximal gradient descent method for optimization. Extensive experiments on 16 real-world data sets with varying degrees of missing annotations validate that our algorithm effectively handles incomplete LDL tasks and outperforms state-of-the-art algorithms.

Keywords:

Label distribution learning

Incomplete annotation

Label correlation

Sparsity

Low-rank structure

Incomplete label distribution learning via label correlation decomposition

Abstract

Journal

Organization