mlsanityData to 7 Oct 2026

Continual Learning for Class- and Domain-Incremental Semantic Segmentation

arXiv 2209.08023 ↗First posted 16 Sep 2022cs.CVcited by 17

Not on the map. Why?

Abstract (arXiv)

The field of continual deep learning is an emerging field and a lot of progress has been made. However, concurrently most of the approaches are only tested on the task of image classification, which is not relevant in the field of intelligent vehicles. Only recently approaches for class-incremental semantic segmentation were proposed. However, all of those approaches are based on some form of knowledge distillation. At the moment there are no investigations on replay-based approaches that are commonly used for object recognition in a continual setting. At the same time while unsupervised domain adaption for semantic segmentation gained a lot of traction, investigations regarding domain-incremental learning in an continual setting is not well-studied. Therefore, the goal of our work is to evaluate and adapt established solutions for continual object recognition to the task of semantic segmentation and to provide baseline methods and evaluation protocols for the task of continual semantic segmentation. We firstly introduce evaluation protocols for the class- and domain-incremental segmentation and analyze selected approaches. We show that the nature of the task of semantic segmentation changes which methods are most effective in mitigating forgetting compared to image classification. Especially, in class-incremental learning knowledge distillation proves to be a vital tool, whereas in domain-incremental learning replay methods are the most effective method.

Key facts. Cited by 17 papers in the map (data through 7 Oct 2026). Builds on 21 of its 26 references that are in the map. First posted on arXiv 16 Sep 2022.

Builds on

21 of 26 references.

References · 21

  1. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  2. 2014Fully Convolutional Networks for Semantic Segmentationcited by 4,845
  3. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  4. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,427
  5. 2016Learning without Forgettingcited by 2,465
  6. 2017Rethinking Atrous Convolution for Semantic Image Segmentationcited by 2,154
  7. 2018BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learningcited by 1,282
  8. 2017Memory Aware Synapses: Learning what (not) to forgetcited by 1,094
  9. 2017CyCADA: Cycle-Consistent Adversarial Domain Adaptationcited by 1,079
  10. 2013An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networkscited by 711
  11. 2018ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentationcited by 623
  12. 2018Re-evaluating Continual Learning Scenarios: A Categorization and Case for Strong Baselinescited by 228
9 more
  1. 2017Incremental Learning of Object Detectors without Catastrophic Forgettingcited by 208
  2. 2020Modeling the Background for Incremental Learning in Semantic Segmentationcited by 189
  3. 2020PLOP: Learning without Forgetting for Continual Semantic Segmentationcited by 177
  4. 2019Incremental Learning Techniques for Semantic Segmentationcited by 141
  5. 2019A continual learning survey: Defying forgetting in classification taskscited by 118
  6. 2020Unsupervised Model Adaptation for Continual Semantic Segmentationcited by 42
  7. 2018Incremental Learning for Semantic Segmentation of Large-Scale Remote Sensing Datacited by 35
  8. 2020Class-Incremental Learning for Semantic Segmentation Re-Using Neither Old Data Nor Old Labelscited by 31
  9. 2020CVPR 2020 Continual Learning in Computer Vision Competition: Approaches, Results, Current Challenges and Future Directionscited by 26

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 3; 2023: 7; 2024: 2; 2025: 4; 2026: 1.

17 papers.

Sort:

  1. 2023A Unified Approach to Domain Incremental Learning with Memory: Theory and Algorithmcited by 32
  2. 2023Survey on Unsupervised Domain Adaptation for Semantic Segmentation for Visual Perception in Automated Drivingcited by 29
  3. 2022Attribution-aware Weight Transfer: A Warm-Start Initialization for Class-Incremental Semantic Segmentationcited by 21
  4. 2023A Survey on Continual Semantic Segmentation: Theory, Challenge, Method and Applicationcited by 21
  5. 2022Learning with Style: Continual Semantic Segmentation Across Tasks and Domainscited by 9
  6. 2022Improving Replay-Based Continual Semantic Segmentation with Smart Data Selectioncited by 8
  7. 2023Generative appearance replay for continual unsupervised domain adaptationcited by 8
  8. 2024Multi-Label Continual Learning for the Medical Domain: A Novel Benchmarkcited by 6
  9. 2025RICO: Two Realistic Benchmarks and an In-Depth Analysis for Incremental Learning in Object Detectioncited by 5
  10. 2023Principles of Forgetting in Domain-Incremental Semantic Segmentation in Adverse Weather Conditionscited by 4
  11. 2025Continual Hyperbolic Learning of Instances and Classescited by 4
  12. 2023Effects of Architectures on Continual Semantic Segmentationcited by 3
5 more
  1. 2024Fairness Evolution in Continual Learning for Medical Imagingcited by 2
  2. 2025Domain-Incremental Semantic Segmentation for Autonomous Driving under Adverse Driving Conditionscited by 2
  3. 2025CLoRA: Parameter-Efficient Continual Learning with Low-Rank Adaptationcited by 2
  4. 2023Continual Cross-Dataset Adaptation in Road Surface Classificationcited by 0
  5. 2026MILE: Mixture of Incremental LoRA Experts for Continual Semantic Segmentation across Domains and Modalitiescited by 0