Abstract:In a real-world scenario, an object could contain multiple tags instead of a single categorical label. To this end, multi-label learning (MLL) emerged. In MLL, the feature distributions are long-tailed and the complex semantic label relation and the long-tailed training samples are the main challenges. Semi-supervised learning is a potential solution. While, existing methods are mainly designed for single class scenario while ignoring the latent label relations. In addition, they cannot well handle the distribution shift commonly existing across source and target domains. To this end, a Semi-supervised Dual Relation Learning (SDRL) framework for multi-label classification is proposed. SDRL utilizes a few labeled samples as well as large scale unlabeled samples in the training stage. It jointly explores the inter-instance feature-level relation and the intra-instance label-level relation even from the unlabeled samples. In our model, a dual-classifier structure is deployed to obtain domain invariant representations. The prediction results from the classifiers are further compared and the most confident predictions are extracted as pseudo labels. A trainable label relation tensor is designed to explicitly explore the pairwise latent label relations and refine the predicted labels. SDRL is able to effectively and efficiently explore the feature-label relation as well as the label-label relation knowledge without any extra semantic knowledge. We evaluated SDRL in general and zero-shot multi-label classification tasks and we concluded that SDRL is superior to other SOTA baselines. Furthermore, extensive ablation studies have been done which reveal the effectiveness of each component in our framework.

Semi-Supervised Multi-Label Linear Discriminant Analysis

Semi-supervised Linear Discriminant Analysis for Dimension Reduction and Classification

Semi-Supervised Multi-Label Dimensionality Reduction Based on Dependence Maximization.

Multi-label Dimensionality Reduction Based on Semi-Supervised Discriminant Analysis

Semi-Supervised Dual Relation Learning for Multi-Label Classification

A soft label based linear discriminant analysis for semi-supervised dimensionality reduction

A Least Squares Formulation of Multi-Label Linear Discriminant Analysis

Multi-label Linear Discriminant Analysis with Locality Consistency.

Saliency-based Weighted Multi-label Linear Discriminant Analysis

Structured Semi-Supervised Discriminant Analysis

Semi-Supervised Multi-Label Dimensionality Reduction Learning by Instance and Label Correlations

Semi-Supervised Discriminant Analysis Via Spectral Transduction.

Disambiguation Enabled Linear Discriminant Analysis for Partial Label Dimensionality Reduction

Semi-supervised Multi-Label Feature Learning Via Label Enlarged Discriminant Analysis.

A Weighted Linear Discriminant Analysis Framework for Multi-Label Feature Extraction

Multimodal Linear Discriminant Analysis Via Structural Sparsity

Dual Relation Semi-Supervised Multi-Label Learning.

Semi-supervised Discriminant Analysis Based on Dependence Estimation

Classification by Semi-Supervised Discriminative Regularization

Research on Label Distribution Dimension Reduction Algorithm Based on MLDA

MBLDA: A Novel Multiple Between-Class Linear Discriminant Analysis