mlsanityData to 7 Oct 2026

Sparse SPN: Depth Completion from Sparse Keypoints

arXiv 2212.00987 ↗First posted 2 Dec 2022cs.CVcited by 3

Not on the map. Why?

Abstract (arXiv)

Our long term goal is to use image-based depth completion to quickly create 3D models from sparse point clouds, e.g. from SfM or SLAM. Much progress has been made in depth completion. However, most current works assume well distributed samples of known depth, e.g. Lidar or random uniform sampling, and perform poorly on uneven samples, such as from keypoints, due to the large unsampled regions. To address this problem, we extend CSPN with multiscale prediction and a dilated kernel, leading to much better completion of keypoint-sampled depth. We also show that a model trained on NYUv2 creates surprisingly good point clouds on ETH3D by completing sparse SfM points.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 26 of its 38 references that are in the map. First posted on arXiv 2 Dec 2022.

Builds on

26 of 38 references.

References · 26

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  4. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  5. 2015Multi-Scale Context Aggregation by Dilated Convolutionscited by 1,128
  6. 2020Non-Local Spatial Propagation Network for Depth Completioncited by 194
  7. 2017Sparse-to-Dense: Depth Prediction from Sparse Depth Samples and a Single Imagecited by 159
  8. 2021PENet: Towards Precise and Efficient Image Guided Depth Completioncited by 146
  9. 2018Learning Depth with Convolutional Spatial Propagation Networkcited by 135
  10. 2018DeepLiDAR: Deep Surface Normal Guided Depth Prediction for Outdoor Scene from Sparse LiDAR Data and Single Color Imagecited by 129
  11. 20223D Common Corruptions and Data Augmentationcited by 120
  12. 2018Deep Depth Completion of a Single RGB-D Imagecited by 110
14 more
  1. 2019CSPN++: Learning Context and Resource Aware Convolutional Spatial Propagation Networks for Depth Completioncited by 110
  2. 2019Depth Completion from Sparse LiDAR Data with Depth-Normal Constraintscited by 108
  3. 2019Multi-Scale Geometric Consistency Guided Multi-View Stereocited by 105
  4. 2019Unsupervised Depth Completion from Visual Inertial Odometrycited by 88
  5. 2017Learning Affinity via Spatial Propagation Networkscited by 79
  6. 2018Sparse and Dense Data with CNNs: Depth Completion and Semantic Segmentationcited by 76
  7. 2022Dynamic Spatial Propagation Network for Depth Completioncited by 70
  8. 2020Adaptive Context-Aware Multi-Modal Network for Depth Completioncited by 69
  9. 2021Unsupervised Depth Completion with Calibrated Backprojection Layerscited by 61
  10. 2021Learning Topology from Synthetic Data for Unsupervised Depth Completioncited by 38
  11. 2020CodeVIO: Visual-Inertial Odometry with Learned Optimizable Dense Depthcited by 23
  12. 2019Deep RGB-D Canonical Correlation Analysis For Sparse Depth Completioncited by 20
  13. 2020Boosting Monocular Depth Estimation with Lightweight 3D Point Fusioncited by 14
  14. 2020Deep Depth Estimation from Visual-Inertial SLAMcited by 13

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 1; 2024: 0; 2025: 1; 2026: 1.

3 papers.

  1. 2025Multi-view Reconstruction via SfM-guided Monocular Depth Estimationcited by 17
  2. 2023SparseDC: Depth Completion from sparse and non-uniform inputscited by 8
  3. 20264Director: Controlling Video World Models with Rigid 3D Geometrycited by 0