# Cascaded Refinement Network for Point Cloud Completion with Self-supervision

arXiv: https://arxiv.org/abs/2010.08719 · HTML: https://mlsanity.com/paper/2010.08719/

## Key facts

- Cited by 20 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Oct 2020.

## Built on by (most cited first, 20 of 20)

- [AdaPoinTr: Diverse Point Cloud Completion with Adaptive Geometry-Aware Transformers](https://mlsanity.com/paper/2301.04545/) (2023)
- [Comprehensive Review of Deep Learning-Based 3D Point Cloud Completion Processing and Analysis](https://mlsanity.com/paper/2203.03311/) (2022)
- [Snowflake Point Deconvolution for Point Cloud Completion and Generation with Skip-Transformer](https://mlsanity.com/paper/2202.09367/) (2022)
- [Cross-modal Learning for Image-Guided Point Cloud Shape Completion](https://mlsanity.com/paper/2209.09552/) (2022)
- [CP3: Unifying Point Cloud Completion by Pretrain-Prompt-Predict Paradigm](https://mlsanity.com/paper/2207.05359/) (2022)
- [Explicitly Guided Information Interaction Network for Cross-modal Point Cloud Completion](https://mlsanity.com/paper/2407.02887/) (2024)
- [PointSea: Point Cloud Completion via Self-structure Augmentation](https://mlsanity.com/paper/2502.17053/) (2025)
- [SCARP: 3D Shape Completion in ARbitrary Poses for Improved Grasping](https://mlsanity.com/paper/2301.07213/) (2023)
- [Lidar Upsampling with Sliced Wasserstein Distance](https://mlsanity.com/paper/2301.13558/) (2023)
- [Implicit and Efficient Point Cloud Completion for 3D Single Object Tracking](https://mlsanity.com/paper/2209.00522/) (2022)
- [3D Shape Completion on Unseen Categories:A Weakly-supervised Approach](https://mlsanity.com/paper/2401.10578/) (2024)
- [FSC: Few-point Shape Completion](https://mlsanity.com/paper/2403.07359/) (2024)
- [Deep learning-based blind image super-resolution with iterative kernel reconstruction and noise estimation](https://mlsanity.com/paper/2404.16564/) (2024)
- [Flexible-weighted Chamfer Distance: Enhanced Objective Function for Point Cloud Completion](https://mlsanity.com/paper/2505.14218/) (2025)
- [Parametric Surface Constrained Upsampler Network for Point Cloud](https://mlsanity.com/paper/2303.08240/) (2023)
- [HeartFormer: Semantic-Aware Dual-Structure Transformers for 3D Four-Chamber Cardiac Point Cloud Reconstruction](https://mlsanity.com/paper/2512.00264/) (2025)
- [PPC-MT: Parallel Point Cloud Completion with Mamba-Transformer Hybrid Architecture](https://mlsanity.com/paper/2603.00870/) (2026)
- [GeoCD: A Differential Local Approximation for Geodesic Chamfer Distance](https://mlsanity.com/paper/2506.23478/) (2025)
- [Multi-modal Multi-task Pre-training for Improved Point Cloud Understanding](https://mlsanity.com/paper/2507.17533/) (2025)
- [MAGE: View-guided Point Cloud Completion with Efficient Modality Alignment and Adaptive Geometry Enhancement](https://mlsanity.com/paper/2607.02568/) (2026)

## Abstract (arXiv)

Point clouds are often sparse and incomplete, which imposes difficulties for real-world applications. Existing shape completion methods tend to generate rough shapes without fine-grained details. Considering this, we introduce a two-branch network for shape completion. The first branch is a cascaded shape completion sub-network to synthesize complete objects, where we propose to use the partial input together with the coarse output to preserve the object details during the dense point reconstruction. The second branch is an auto-encoder to reconstruct the original partial input. The two branches share a same feature extractor to learn an accurate global feature for shape completion. Furthermore, we propose two strategies to enable the training of our network when ground truth data are not available. This is to mitigate the dependence of existing approaches on large amounts of ground truth training data that are often difficult to obtain in real-world applications. Additionally, our proposed strategies are also able to improve the reconstruction quality for fully supervised learning. We verify our approach in self-supervised, semi-supervised and fully supervised settings with superior performances. Quantitative and qualitative results on different datasets demonstrate that our method achieves more realistic outputs than state-of-the-art approaches on the point cloud completion task.