mlsanityData to 14 Oct 2026

SCoDA: Domain Adaptive Shape Completion for Real Scans

arXiv 2304.10179 ↗First posted 20 Apr 2023cs.CVcited by 9

Not on the map. Why?

Abstract (arXiv)

3D shape completion from point clouds is a challenging task, especially from scans of real-world objects. Considering the paucity of 3D shape ground truths for real scans, existing works mainly focus on benchmarking this task on synthetic data, e.g. 3D computer-aided design models. However, the domain gap between synthetic and real data limits the generalizability of these methods. Thus, we propose a new task, SCoDA, for the domain adaptation of real scan shape completion from synthetic data. A new dataset, ScanSalon, is contributed with a bunch of elaborate 3D models created by skillful artists according to scans. To address this new task, we propose a novel cross-domain feature fusion method for knowledge transfer and a novel volume-consistent self-training framework for robust learning from real data. Extensive experiments prove our method is effective to bring an improvement of 6%~7% mIoU.

Key facts. Cited by 9 papers in the map (data through 14 Oct 2026). Builds on 92 of its 105 references that are in the map. First posted on arXiv 20 Apr 2023.

Builds on

92 of 105 references.

References · 92

  1. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,117
  2. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,897
  3. 2019DeepSDF: Learning Continuous Signed Distance Functions for Shape Representationcited by 2,689
  4. 2020FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidencecited by 2,085
  5. 2018Occupancy Networks: Learning 3D Reconstruction in Function Spacecited by 2,041
  6. 2017Adversarial Discriminative Domain Adaptationcited by 1,311
  7. 2018Learning Implicit Fields for Generative Shape Modelingcited by 1,060
  8. 2019Unsupervised Data Augmentation for Consistency Trainingcited by 1,060
  9. 2019PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human Digitizationcited by 818
  10. 2020Convolutional Occupancy Networkscited by 761
  11. 2017Maximum Classifier Discrepancy for Unsupervised Domain Adaptationcited by 683
  12. 2016Deep Transfer Learning with Joint Adaptation Networkscited by 612
38 more
  1. 20163D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstructioncited by 588
  2. 2018AtlasNet: A Papier-Mach'e Approach to Learning 3D Surface Generationcited by 539
  3. 2018PCN: Point Completion Networkcited by 384
  4. 2020Local Implicit Grid Representations for 3D Scenescited by 373
  5. 2020Implicit Functions in Feature Space for 3D Shape Reconstruction and Completioncited by 343
  6. 2019DISN: Deep Implicit Surface Network for High-quality Single-view 3D Reconstructioncited by 333
  7. 2021DAFormer: Improving Network Architectures and Training Strategies for Domain-Adaptive Semantic Segmentationcited by 300
  8. 2021PoinTr: Diverse Point Cloud Completion with Geometry-Aware Transformerscited by 286
  9. 2019What Do Single-view 3D Reconstruction Networks Learn?cited by 260
  10. 2018SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloudcited by 228
  11. 2018DeepJDOT: Deep Joint Distribution Optimal Transport for Unsupervised Domain Adaptationcited by 214
  12. 2020MetaSDF: Meta-learning Signed Distance Functionscited by 214
  13. 2022CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D Point Cloud Understandingcited by 198
  14. 2021Shape As Points: A Differentiable Poisson Solvercited by 187
  15. 2020Instance Adaptive Self-Training for Unsupervised Domain Adaptationcited by 177
  16. 2020GRNet: Gridding Residual Network for Dense Point Cloud Completioncited by 174
  17. 2019Pix2Vox: Context-aware 3D Reconstruction from Single and Multi-view Imagescited by 164
  18. 2021SnowflakeNet: Point Cloud Completion by Snowflake Point Deconvolution with Skip-Transformercited by 164
  19. 2020PF-Net: Point Fractal Network for 3D Point Cloud Completioncited by 146
  20. 2018Scan2CAD: Learning CAD Model Alignment in RGB-D Scanscited by 142
  21. 2017Learning Efficient Point Cloud Generation for Dense 3D Object Reconstructioncited by 133
  22. 2019Morphing and Sampling Network for Dense Point Cloud Completioncited by 132
  23. 2020Gradually Vanishing Bridge for Adversarial Domain Adaptationcited by 130
  24. 2019xMUDA: Cross-Modal Unsupervised Domain Adaptation for 3D Semantic Segmentationcited by 129
  25. 2020Harmonizing Transferability and Discriminability for Adapting Object Detectorscited by 126
  26. 2020Points2Surf: Learning Implicit Surfaces from Point Cloud Patchescited by 122
  27. 2020Complete & Label: A Domain Adaptation Approach to Semantic Segmentation of LiDAR Point Cloudscited by 116
  28. 2019PointDAN: A Multi-Scale 3D Domain Adaption Network for Point Cloud Representationcited by 114
  29. 2020Cascaded Refinement Network for Point Cloud Completioncited by 112
  30. 2018Deep Geometric Prior for Surface Reconstructioncited by 110
  31. 2019A Robust Learning Approach to Domain Adaptive Object Detectioncited by 109
  32. 2017MarrNet: 3D Shape Reconstruction via 2.5D Sketchescited by 108
  33. 2020CrDoCo: Pixel-level Domain Transfer with Cross-Domain Consistencycited by 107
  34. 2020Point Cloud Completion by Skip-attention Network with Hierarchical Foldingcited by 106
  35. 2021Variational Relational Point Completion Networkcited by 104
  36. 2021AdaMatch: A Unified Approach to Semi-Supervised Learning and Domain Adaptationcited by 104
  37. 2020Self-Supervised Learning for Domain Adaptation on Point-Cloudscited by 99
  38. 2019Learning elementary structures for 3D shape generation and matchingcited by 98

50 most cited of 92 All 92 (JSON)

Built on by

Cited by year

Citing papers per year: 2023: 2; 2024: 5; 2025: 2.

9 papers.

  1. 2024PCoTTA: Continual Test-Time Adaptation for Multi-Task Point Cloud Understandingcited by 11
  2. 2025GenPC: Zero-shot Point Cloud Completion via 3D Generative Priorscited by 9
  3. 2024DreamDissector: Learning Disentangled Text-to-3D Generation from 2D Diffusion Priorscited by 8
  4. 20243D Shape Completion on Unseen Categories:A Weakly-supervised Approachcited by 6
  5. 2023LASA: Instance Reconstruction from Real Scans using A Large-scale Aligned Shape Annotation Datasetcited by 5
  6. 2024Deep Learning for 3D Point Cloud Enhancement: A Surveycited by 4
  7. 2024Self-supervised Shape Completion via Involution and Implicit Correspondencescited by 2
  8. 2025Learning Generalizable Shape Completion with SIM(3) Equivariancecited by 2
  9. 2023Universal Semi-supervised Model Adaptation via Collaborative Consistency Trainingcited by 1