mlsanityData to 7 Oct 2026

Pseudo-label Learning with Calibrated Confidence Using an Energy-based Model

arXiv 2404.09585 ↗First posted 15 Apr 2024cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

In pseudo-labeling (PL), which is a type of semi-supervised learning, pseudo-labels are assigned based on the confidence scores provided by the classifier; therefore, accurate confidence is important for successful PL. In this study, we propose a PL algorithm based on an energy-based model (EBM), which is referred to as the energy-based PL (EBPL). In EBPL, a neural network-based classifier and an EBM are jointly trained by sharing their feature extraction parts. This approach enables the model to learn both the class decision boundary and input data distribution, enhancing confidence calibration during network training. The experimental results demonstrate that EBPL outperforms the existing PL method in semi-supervised image classification tasks, with superior confidence calibration error and recognition accuracy.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 24 of its 43 references that are in the map. First posted on arXiv 15 Apr 2024.

Builds on

24 of 43 references.

References · 24

  1. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
  2. 2017On Calibration of Modern Neural Networkscited by 4,860
  3. 2017Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithmscited by 4,283
  4. 2016Wide Residual Networkscited by 3,057
  5. 2020FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidencecited by 2,080
  6. 2019Self-training with Noisy Student improves ImageNet classificationcited by 1,304
  7. 2021MedMNIST v2 -- A large-scale lightweight benchmark for 2D and 3D biomedical image classificationcited by 682
  8. 2019Pseudo-Labeling and Confirmation Bias in Deep Semi-Supervised Learningcited by 507
  9. 2019Your Classifier is Secretly an Energy Based Model and You Should Treat it Like Onecited by 402
  10. 2020Meta Pseudo Labelscited by 372
  11. 2019On Mixup Training: Improved Calibration and Predictive Uncertainty for Deep Neural Networkscited by 304
  12. 2021In Defense of Pseudo-Labeling: An Uncertainty-Aware Pseudo-label Selection Framework for Semi-Supervised Learningcited by 301
12 more
  1. 2020Improved Contrastive Divergence Training of Energy Based Modelscited by 127
  2. 2020Residual Energy-Based Models for Text Generationcited by 95
  3. 2021Be Confident! Towards Trustworthy Graph Neural Networks via Confidence Calibrationcited by 91
  4. 2022Debiased Learning from Naturally Imbalanced Pseudo-Labelscited by 85
  5. 2022BoostMIS: Boosting Medical Image Semi-supervised Learning with Adaptive Pseudo Labeling and Informative Active Annotationcited by 54
  6. 2018Divergence Triangle for Joint Training of Generator Model, Energy-based Model, and Inference Modelcited by 34
  7. 2018Learning for Single-Shot Confidence Calibration in Deep Neural Networks through Stochastic Inferencescited by 25
  8. 2020EDropout: Energy-Based Dropout and Pruning of Deep Neural Networkscited by 20
  9. 2022Calibrating Deep Neural Networks using Explicit Regularisation and Dynamic Data Pruningcited by 18
  10. 2019Energy-Based Continuous Inverse Optimal Controlcited by 14
  11. 2023Bridging Precision and Confidence: A Train-Time Loss for Calibrating Object Detectioncited by 12
  12. 2023A Hybrid of Generative and Discriminative Models Based on the Gaussian-coupled Softmax Layercited by 4

Built on by

Not cited yet.