mlsanityData to 7 Oct 2026

Energy-constrained Self-training for Unsupervised Domain Adaptation

arXiv 2101.00316 ↗First posted 1 Jan 2021cs.CV cs.AI cs.LGcited by 16

Not on the map. Why?

Abstract (arXiv)

Unsupervised domain adaptation (UDA) aims to transfer the knowledge on a labeled source domain distribution to perform well on an unlabeled target domain. Recently, the deep self-training involves an iterative process of predicting on the target domain and then taking the confident predictions as hard pseudo-labels for retraining. However, the pseudo-labels are usually unreliable, and easily leading to deviated solutions with propagated errors. In this paper, we resort to the energy-based model and constrain the training of the unlabeled target sample with the energy function minimization objective. It can be applied as a simple additional regularization. In this framework, it is possible to gain the benefits of the energy-based model, while retaining strong discriminative performance following a plug-and-play fashion. We deliver extensive experiments on the most popular and large scale UDA benchmarks of image classification as well as semantic segmentation to demonstrate its generality and effectiveness.

Key facts. Cited by 16 papers in the map (data through 7 Oct 2026). Builds on 33 of its 56 references that are in the map. First posted on arXiv 1 Jan 2021.

Builds on

33 of 56 references.

References · 33

  1. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,427
  2. 2015Domain-Adversarial Training of Neural Networkscited by 3,023
  3. 2016Pyramid Scene Parsing Networkcited by 2,472
  4. 2015Learning Transferable Features with Deep Adaptation Networkscited by 1,323
  5. 2017CyCADA: Cycle-Consistent Adversarial Domain Adaptationcited by 1,079
  6. 2016Playing for Data: Ground Truth from Computer Gamescited by 816
  7. 2017Maximum Classifier Discrepancy for Unsupervised Domain Adaptationcited by 682
  8. 2018ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentationcited by 623
  9. 2018Learning to Adapt Structured Output Space for Semantic Segmentationcited by 584
  10. 2019Confidence Regularized Self-Trainingcited by 418
  11. 2019Your Classifier is Secretly an Energy Based Model and You Should Treat it Like Onecited by 402
  12. 2017Generate To Adapt: Aligning Domains using Generative Adversarial Networkscited by 183
21 more
  1. 2016A Theory of Generative ConvNetcited by 174
  2. 2019Domain Adaptation for Structured Output via Discriminative Patch Representationscited by 150
  3. 2018An introduction to domain adaptation and transfer learningcited by 112
  4. 2017Adversarial Dropout Regularizationcited by 106
  5. 2019Constructing Self-motivated Pyramid Curriculums for Cross-Domain Semantic Segmentation: A Non-Adversarial Approachcited by 103
  6. 2018DCAN: Dual Channel-wise Alignment Networks for Unsupervised Scene Adaptationcited by 101
  7. 2018Fully Convolutional Adaptation Networks for Semantic Segmentationcited by 75
  8. 2017Unsupervised Domain Adaptation with Similarity Learningcited by 73
  9. 2018Label Refinery: Improving ImageNet Classification through Label Progressioncited by 65
  10. 2019Open Set Domain Adaptation for Image and Action Recognitioncited by 30
  11. 2020Importance-Aware Semantic Segmentation in Self-Driving with Discrete Wasserstein Trainingcited by 27
  12. 2019Deep Verifier Networks: Verification of Deep Discriminative Models with Deep Generative Modelscited by 25
  13. 2019Unsupervised Domain Adaptation via Calibrating Uncertaintiescited by 15
  14. 2020AUTO3D: Novel view synthesis through unsupervisely learned variational viewpoint and global 3D representationcited by 14
  15. 2019Unimodal-uniform Constrained Wasserstein Training for Medical Diagnosiscited by 13
  16. 2019Dependency-aware Attention Control for Unconstrained Face Recognition with Image Setscited by 11
  17. 2019Conservative Wasserstein Training for Pose Estimationcited by 11
  18. 2019Permutation-invariant Feature Restructuring for Correlation-aware Image Set-based Recognitioncited by 10
  19. 2018Generative Modeling by Inclusive Neural Random Fields with Applications in Image Generation and Anomaly Detectioncited by 9
  20. 2019Task-Discriminative Domain Alignment for Unsupervised Domain Adaptationcited by 5
  21. 2019Attention Control with Metric Learning Alignment for Image Set-based Recognitioncited by 2

Built on by

Cited by year

Citing papers per year: 2021: 5; 2022: 6; 2023: 2; 2024: 2; 2025: 1.

16 papers.

Sort:

  1. 2022Deep Unsupervised Domain Adaptation: A Review of Recent Advances and Perspectivescited by 98
  2. 2021Adversarial Unsupervised Domain Adaptation with Conditional and Label Shift: Infer, Align and Iteratecited by 43
  3. 2021Adapting Off-the-Shelf Source Segmenter for Target Medical Image Segmentationcited by 39
  4. 2023Test-Time Training on Video Streamscited by 37
  5. 2021Generative Self-training for Cross-domain Unsupervised Tagged-to-Cine MRI Synthesiscited by 19
  6. 2022Memory Consistent Unsupervised Off-the-Shelf Model Adaptation for Source-Relaxed Medical Image Segmentationcited by 15
  7. 2021Recursively Conditional Gaussian for Ordinal Unsupervised Domain Adaptationcited by 10
  8. 2022Unsupervised Domain Adaptation for Segmentation with Black-box Source Modelcited by 7
  9. 2024Resolving Domain Shift For Representations Of Speech In Non-Invasive Brain Recordingscited by 5
  10. 2021Interpreting Depression From Question-wise Long-term Video Recording of SDS Evaluationcited by 4
  11. 2022Variational Inference for Quantifying Inter-observer Variability in Segmentation of Anatomical Structurescited by 4
  12. 2022Subtype-Aware Dynamic Unsupervised Domain Adaptationcited by 2
4 more
  1. 2022Constraining Pseudo-label in Self-training Unsupervised Domain Adaptation with Energy-based Modelcited by 2
  2. 2023Attentive Continuous Generative Self-training for Unsupervised Domain Adaptive Medical Image Translationcited by 2
  3. 2024Domain Adaptation Using Pseudo Labelscited by 1
  4. 2025TCSA-UDA: Text-Driven Cross-Semantic Alignment for Unsupervised Domain Adaptation in Medical Image Segmentationcited by 0