Image-Feature Weak-to-Strong Consistency: An Enhanced Paradigm for Semi-Supervised Learning

TL;DR

This paper proposes a novel semi-supervised learning paradigm that combines image and feature perturbations with a triple-branch structure and confidence-based sample identification to improve performance across various benchmarks.

Contribution

It introduces a feature-level perturbation approach and a triple-branch structure for enhanced semi-supervised learning, along with a confidence-based strategy to handle naive samples.

Findings

Significant performance improvements on multiple benchmarks.

Effective handling of naive and challenging samples.

Compatibility with existing SSL algorithms.

Abstract

Image-level weak-to-strong consistency serves as the predominant paradigm in semi-supervised learning~(SSL) due to its simplicity and impressive performance. Nonetheless, this approach confines all perturbations to the image level and suffers from the excessive presence of naive samples, thus necessitating further improvement. In this paper, we introduce feature-level perturbation with varying intensities and forms to expand the augmentation space, establishing the image-feature weak-to-strong consistency paradigm. Furthermore, our paradigm develops a triple-branch structure, which facilitates interactions between both types of perturbations within one branch to boost their synergy. Additionally, we present a confidence-based identification strategy to distinguish between naive and challenging samples, thus introducing additional challenges exclusively for naive samples. Notably, our paradigm can seamlessly integrate with existing SSL methods. We apply the proposed paradigm to several representative algorithms and conduct experiments on multiple benchmarks, including both balanced and imbalanced distributions for labeled samples. The results demonstrate a significant enhancement in the performance of existing SSL algorithms.