An Overview of Deep Semi-Supervised Learning

Abstract

Deep neural networks demonstrated their ability to provide remarkable performances on a wide range of supervised learning tasks (e.g., image classification) when trained on extensive collections of labeled data (e.g., ImageNet). However, creating such large datasets requires a considerable amount of resources, time, and effort. Such resources may not be available in many practical cases, limiting the adoption and the application of many deep learning methods. In a search for more data-efficient deep learning methods to overcome the need for large annotated datasets, there is a rising research interest in semi-supervised learning and its applications to deep neural networks to reduce the amount of labeled data required, by either developing novel methods or adopting existing semi-supervised learning frameworks for a deep learning setting. In this paper, we provide a comprehensive overview of deep semi-supervised learning, starting with an introduction to the field, followed by a summarization of the dominant semi-supervised approaches in deep learning.

Preprint

References in corpus (13)

• Semi-Supervised Classification with Graph Convolutional Networks
• Semi-Supervised Learning with Deep Generative Models
• Energy-based Generative Adversarial Network
• Variational Graph Auto-Encoders
• Dual Learning for Machine Translation
• AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty
• Billion-scale semi-supervised learning for image classification
• Self-supervised Visual Feature Learning with Deep Neural Networks: A Survey
• ReMixMatch: Semi-Supervised Learning with Distribution Alignment and Augmentation Anchoring
• HARP: Hierarchical Representation Learning for Networks
• Co-training for Demographic Classification Using Deep Learning from Label Proportions
• Consistency Regularization for Generative Adversarial Networks
• Domain Adaptations for Computer Vision Applications