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.

High Reliable Multi-View Semi-Supervised Learning with Extremely Sparse Labeled Data

Active Learning with Extremely Sparse Labeled Examples.

Semi-supervised learning with very few labeled training examples

Dual-Classifier Collaborative Method Based on Semi-Supervised Active Learning

Semi-supervised learning with extremely sparse labeled data on multiple semi-supervised assumptions

Multiview Semi-supervised Learning

Semi-Supervised Dual Relation Learning for Multi-Label Classification

Embedding Regularizer Learning for Multi-View Semi-Supervised Classification.

Sparse Semi-supervised Learning Using Conjugate Functions

Unlabeled Data and Multiple Views

Multiple-View Multiple-Learner Semi-Supervised Learning

Semi-supervised Deep Multi-view Stereo

Deep Incomplete Multi-View Learning Network with Insufficient Label Information

Dlsa: Semi-supervised Partial Label Learning Via Dependence-Maximized Label Set Assignment

Multiview Supervised Learning

Semantic Features for Multi-view Semi-supervised and Active Learning of Text Classification

Sparse low-redundancy multilabel feature selection based on dynamic local structure preservation and triple graphs exploration

Multi-View Semi-Supervised Learning For Web Image Annotation

Latent Multi-view Semi-Supervised Classification.

Semi-Supervised Multi-Label Linear Discriminant Analysis

Large-Margin Predictive Latent Subspace Learning for Multiview Data Analysis.