mlsanityData to 13 Oct 2026

Balanced Residual Distillation Learning for 3D Point Cloud Class-Incremental Semantic Segmentation

arXiv 2408.01356 ↗First posted 2 Aug 2024cs.CVcited by 6

Not on the map. Why?

Abstract (arXiv)

Class-incremental learning (CIL) enables continuous learning of new classes while mitigating catastrophic forgetting of old ones. For the performance breakthrough of CIL, it is essential yet challenging to effectively refine past knowledge from the base model and balance it with new learning. However, such a challenge has not been considered in current research. This work proposes a balanced residual distillation learning framework (BRDL) to address this gap and advance CIL performance. BRDL introduces a residual distillation strategy to dynamically refine past knowledge by expanding the network structure and a balanced pseudo-label learning strategy to mitigate class bias and balance learning between old and new classes. We apply the proposed BRDL to a challenging 3D point cloud semantic segmentation task where the data is unordered and unstructured. Extensive experimental results demonstrate that BRDL sets a new benchmark with an outstanding balance capability in class-biased scenarios.

Key facts. Cited by 6 papers in the map (data through 13 Oct 2026). Builds on 25 of its 41 references that are in the map. First posted on arXiv 2 Aug 2024.

Builds on

25 of 41 references.

References · 25

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,093
  2. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,231
  3. 2016Overcoming catastrophic forgetting in neural networkscited by 4,753
  4. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,389
  5. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,096
  6. 2016Learning without Forgettingcited by 2,479
  7. 2016iCaRL: Incremental Classifier and Representation Learningcited by 2,323
  8. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,222
  9. 2018Experience Replay for Continual Learningcited by 1,082
  10. 2017Lifelong Learning with Dynamically Expandable Networkscited by 692
  11. 2018End-to-End Incremental Learningcited by 597
  12. 2021DER: Dynamically Expandable Representation for Class Incremental Learningcited by 355
13 more
  1. 2019Maintaining Discrimination and Fairness in Class Incremental Learningcited by 281
  2. 2022FOSTER: Feature Boosting and Compression for Class-Incremental Learningcited by 228
  3. 2019Compacting, Picking and Growing for Unforgetting Continual Learningcited by 207
  4. 2022Stratified Transformer for 3D Point Cloud Segmentationcited by 166
  5. 2019Uncertainty-based Continual Learning with Adaptive Regularizationcited by 144
  6. 2022Class-Incremental Learning by Knowledge Distillation with Adaptive Feature Consolidationcited by 113
  7. 2018A LiDAR Point Cloud Generator: from a Virtual World to Autonomous Drivingcited by 54
  8. 2022Uncertainty-aware Contrastive Distillation for Incremental Semantic Segmentationcited by 37
  9. 2020Self-Training for Class-Incremental Semantic Segmentationcited by 32
  10. 2023Hierarchical Point-based Active Learning for Semi-supervised Point Cloud Semantic Segmentationcited by 21
  11. 2020I3DOL: Incremental 3D Object Learning without Catastrophic Forgettingcited by 20
  12. 2023NAPA-VQ: Neighborhood Aware Prototype Augmentation with Vector Quantization for Continual Learningcited by 12
  13. 2022LACV-Net: Semantic Segmentation of Large-Scale Point Cloud Scene via Local Adaptive and Comprehensive VLADcited by 5

Built on by

Cited by year (2026 to 13 Oct)

Citing papers per year: 2025: 4; 2026: 2.

6 papers.

  1. 2025SRKD: Towards Efficient 3D Point Cloud Segmentation via Structure- and Relation-aware Knowledge Distillationcited by 9
  2. 2025Decoupling Continual Semantic Segmentationcited by 2
  3. 2025CLIMB-3D: Continual Learning for Imbalanced 3D Instance Segmentationcited by 1
  4. 2025FDBPL: Faster Distillation-Based Prompt Learning for Region-Aware Vision-Language Models Adaptationcited by 1
  5. 2026SCOPE: Scene-Contextualized Incremental Few-Shot 3D Segmentationcited by 1
  6. 2026Joint Optimization of Storage and Loading for High-Performance 3D Point Cloud Data Processingcited by 0