Energy Models for Better Pseudo-Labels: Improving Semi-Supervised Classification with the 1-Laplacian Graph Energy

Abstract

Semi-supervised classification is a great focus of interest, as in real-world scenarios obtaining labels is expensive, time-consuming and might require expert knowledge. This has motivated the fast development of semi-supervised techniques, whose performance is on a par with or better than supervised approaches. A current major challenge for semi-supervised techniques is how to better handle the network calibration and confirmation bias problems for improving performance. In this work, we argue that energy models are an effective alternative to such problems. With this motivation in mind, we propose a hybrid framework for semi-supervised classification called CREPE model (1-Lapla$\mathbf{C}$ian g$\mathbf{R}$aph $\mathbf{E}$nergy for $\mathbf{P}$seudo-lab$\mathbf{E}$ls). Firstly, we introduce a new energy model based on the non-smooth ${\ell }_1$ norm of the normalised graph 1-Laplacian. Our functional enforces a sufficiently smooth solution and strengthens the intrinsic relation between the labelled and unlabelled data. Secondly, we provide a theoretical analysis for our proposed scheme and show that the solution trajectory does converge to a non-constant steady point. Thirdly, we derive the connection of our energy model for pseudo-labelling. We show that our energy model produces more meaningful pseudo-labels than the ones generated directly by a deep network. We extensively evaluate our framework, through numerical and visual experiments, using six benchmarking datasets for natural and medical images. We demonstrate that our technique reports state-of-the-art results for semi-supervised classification.

References in corpus (17)

• Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
• Improved Baselines with Momentum Contrastive Learning
• On Calibration of Modern Neural Networks
• Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results
• Interpolation Consistency Training for Semi-Supervised Learning
• AutoAugment: Learning Augmentation Policies from Data
• Semi-Supervised Learning with Ladder Networks
• There Are Many Consistent Explanations of Unlabeled Data: Why You Should Average
• An Inverse Power Method for Nonlinear Eigenproblems with Applications in 1-Spectral Clustering and Sparse PCA
• Weakly Supervised Medical Diagnosis and Localization from Multiple Resolutions
• In Defense of Pseudo-Labeling: An Uncertainty-Aware Pseudo-label Selection Framework for Semi-Supervised Learning
• The Total Variation on Hypergraphs - Learning on Hypergraphs Revisited
• GraphX${}^{NET}-$ Chest X-Ray Classification Under Extreme Minimal Supervision
• Poisson Learning: Graph Based Semi-Supervised Learning At Very Low Label Rates
• An Adaptive Total Variation Algorithm for Computing the Balanced Cut of a Graph
• Tight Continuous Relaxation of the Balanced $k$-Cut Problem
• Self-supervised Mean Teacher for Semi-supervised Chest X-ray Classification

Cited by in corpus (4)

• On the Duality between Network Flows and Network Lasso
• Nonlinear Power Method for Computing Eigenvectors of Proximal Operators and Neural Networks
• Revealing Stable and Unstable Modes of Generic Denoisers through Nonlinear Eigenvalue Analysis
• CycleCluster: Modernising Clustering Regularisation for Deep Semi-Supervised Classification