mlsanityData to 14 Oct 2026

ProtoDepth: Unsupervised Continual Depth Completion with Prototypes

arXiv 2503.12745 ↗First posted 17 Mar 2025cs.CVcited by 11

Not on the map. Why?

Abstract (arXiv)

We present ProtoDepth, a novel prototype-based approach for continual learning of unsupervised depth completion, the multimodal 3D reconstruction task of predicting dense depth maps from RGB images and sparse point clouds. The unsupervised learning paradigm is well-suited for continual learning, as ground truth is not needed. However, when training on new non-stationary distributions, depth completion models will catastrophically forget previously learned information. We address forgetting by learning prototype sets that adapt the latent features of a frozen pretrained model to new domains. Since the original weights are not modified, ProtoDepth does not forget when test-time domain identity is known. To extend ProtoDepth to the challenging setting where the test-time domain identity is withheld, we propose to learn domain descriptors that enable the model to select the appropriate prototype set for inference. We evaluate ProtoDepth on benchmark dataset sequences, where we reduce forgetting compared to baselines by 52.2% for indoor and 53.2% for outdoor to achieve the state of the art.

Key facts. Cited by 11 papers in the map (data through 14 Oct 2026). Builds on 100 of its 124 references that are in the map. First posted on arXiv 17 Mar 2025.

Builds on

100 of 124 references.

References · 100

  1. 2017Attention Is All You Needcited by 60,361
  2. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  3. 2016Overcoming catastrophic forgetting in neural networkscited by 4,787
  4. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,117
  5. 2017Prototypical Networks for Few-shot Learningcited by 3,015
  6. 2020Unsupervised Learning of Visual Features by Contrasting Cluster Assignmentscited by 2,857
  7. 2016Learning without Forgettingcited by 2,496
  8. 2019Scalability in Perception for Autonomous Driving: Waymo Open Datasetcited by 2,451
  9. 2016iCaRL: Incremental Classifier and Representation Learningcited by 2,333
  10. 2017Gradient Episodic Memory for Continual Learningcited by 1,898
  11. 2014FitNets: Hints for Thin Deep Netscited by 1,653
  12. 2016Progressive Neural Networkscited by 1,361
38 more
  1. 2024Depth Anything: Unleashing the Power of Large-Scale Unlabeled Datacited by 1,360
  2. 2017Continual Learning with Deep Generative Replaycited by 1,158
  3. 2017Memory Aware Synapses: Learning what (not) to forgetcited by 1,105
  4. 2018Experience Replay for Continual Learningcited by 1,095
  5. 2021Learning to Prompt for Continual Learningcited by 944
  6. 2020Dark Experience for General Continual Learning: a Strong, Simple Baselinecited by 911
  7. 2017PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruningcited by 898
  8. 2021Uformer: A General U-Shaped Transformer for Image Restorationcited by 866
  9. 2018Riemannian Walk for Incremental Learning: Understanding Forgetting and Intransigencecited by 742
  10. 2018This Looks Like That: Deep Learning for Interpretable Image Recognitioncited by 737
  11. 2019Large Scale Incremental Learningcited by 732
  12. 2017Lifelong Learning with Dynamically Expandable Networkscited by 694
  13. 2018Overcoming catastrophic forgetting with hard attention to the taskcited by 681
  14. 2022DualPrompt: Complementary Prompting for Rehearsal-free Continual Learningcited by 642
  15. 2020Prototypical Contrastive Learning of Unsupervised Representationscited by 640
  16. 2018Learning to Learn without Forgetting by Maximizing Transfer and Minimizing Interferencecited by 521
  17. 2022CODA-Prompt: COntinual Decomposed Attention-based Prompting for Rehearsal-Free Continual Learningcited by 450
  18. 2024Depth Pro: Sharp Monocular Metric Depth in Less Than a Secondcited by 404
  19. 2018Dynamic Few-Shot Visual Learning without Forgettingcited by 385
  20. 2016Virtual Worlds as Proxy for Multi-Object Tracking Analysiscited by 341
  21. 2017Variational Continual Learningcited by 339
  22. 2017Sparsity Invariant CNNscited by 326
  23. 2017Few-Shot Learning with Graph Neural Networkscited by 324
  24. 2019Learn to Grow: A Continual Structure Learning Framework for Overcoming Catastrophic Forgettingcited by 302
  25. 2021DyTox: Transformers for Continual Learning with DYnamic TOken eXpansioncited by 291
  26. 2021Online Continual Learning in Image Classification: An Empirical Surveycited by 245
  27. 2020Supermasks in Superpositioncited by 227
  28. 2020Prototype Mixture Models for Few-shot Semantic Segmentationcited by 190
  29. 2019Learning to Remember: A Synaptic Plasticity Driven Framework for Continual Learningcited by 182
  30. 2017FearNet: Brain-Inspired Model for Incremental Learningcited by 177
  31. 2020Using Hindsight to Anchor Past Knowledge in Continual Learningcited by 161
  32. 2019Continual Unsupervised Representation Learningcited by 147
  33. 2020Continual Prototype Evolution: Learning Online from Non-Stationary Data Streamscited by 144
  34. 2018Self-supervised Sparse-to-Dense: Self-supervised Depth Completion from LiDAR and Monocular Cameracited by 140
  35. 2018Memory Efficient Experience Replay for Streaming Learningcited by 128
  36. 2019Functional Regularisation for Continual Learning with Gaussian Processescited by 123
  37. 2022Self-Sustaining Representation Expansion for Non-Exemplar Class-Incremental Learningcited by 118
  38. 2019Overcoming Catastrophic Forgetting with Unlabeled Data in the Wildcited by 111

50 most cited of 100 All 100 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 4; 2026: 7.

11 papers.

  1. 2025Extending Foundational Monocular Depth Estimators to Fisheye Cameras with Calibration Tokenscited by 9
  2. 2025Radar-Guided Polynomial Fitting for Metric Depth Estimationcited by 5
  3. 2025ETA: Energy-based Test-time Adaptation for Depth Completioncited by 4
  4. 2026SHOW3D: Capturing Scenes of 3D Hands and Objects in the Wildcited by 4
  5. 2026Depth Completion as Parameter-Efficient Test-Time Adaptationcited by 3
  6. 2026StreamTTO: Efficient Online Test-Time Optimization for Video Depth Completioncited by 3
  7. 2026VisTa3D: A Dataset and Benchmark for Thin Object Reconstruction from Vision, Tactile, and 3D Point Cloudscited by 1
  8. 2025Geometry-Aware Sparse Depth Sampling for High-Fidelity RGB-D Depth Completion in Robotic Systemscited by 0
  9. 2026Beyond Prompt Degradation: Prototype-guided Dual-pool Prompting for Incremental Object Detectioncited by 0
  10. 2026Unsupervised Continual Clustering via Forward-Backward Knowledge Distillationcited by 0
  11. 2026From Perspective to Fisheye Depth Estimation and Open-Vocabulary Segmentationcited by 0