mlsanityData to 14 Oct 2026

Preserving Guidance in Cost-Volume Retrieval under Extreme LiDAR Sparsity in Iterative Stereo

arXiv 2507.19738 ↗First posted 26 Jul 2025cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

While accurate LiDAR depth has been shown to improve stereo matching, high-end LiDAR remains costly and difficult to deploy at scale, motivating guidance from sparse LiDAR measurements. In this paper, we revisit how extremely sparse LiDAR can guide iterative stereo by examining its cost-volume retrieval mechanism. Our analysis shows that LiDAR guidance degrades sharply when only a few hundred points are available, and we provide a signal-processing explanation for why iterative stereo fails to fully exploit such sparse input. This insight leads to a simple yet highly effective remedy: pre-filling the initial disparity map so that cost-volume retrieval becomes more reliable under extremely sparse LiDAR. We find that pre-filling is also effective when injecting LiDAR depth into image features via early fusion, but for a fundamentally different reason, necessitating a distinct pre-filling approach. Combining both forms of guidance, our proposed Guided RAFT-Stereo (GRAFT-Stereo) achieves substantial improvements over existing LiDAR-guided stereo methods across multiple datasets.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 48 of its 65 references that are in the map. First posted on arXiv 26 Jul 2025.

Builds on

48 of 65 references.

References · 48

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  3. 2020RAFT: Recurrent All-Pairs Field Transforms for Optical Flowcited by 2,415
  4. 2018Pyramid Stereo Matching Networkcited by 489
  5. 2017End-to-End Learning of Geometry and Context for Deep Stereo Regressioncited by 360
  6. 2020Virtual KITTI 2cited by 353
  7. 2021RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matchingcited by 353
  8. 2017Sparsity Invariant CNNscited by 326
  9. 2022Practical Stereo Matching via Cascaded Recurrent Network with Adaptive Correlationcited by 197
  10. 2020Non-Local Spatial Propagation Network for Depth Completioncited by 195
  11. 2020Revisiting Stereo Depth Estimation From a Sequence-to-Sequence Perspective with Transformerscited by 190
  12. 2023Iterative Geometry Encoding Volume for Stereo Matchingcited by 190
36 more
  1. 2019Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Drivingcited by 188
  2. 2024SEA-RAFT: Simple, Efficient, Accurate RAFT for Optical Flowcited by 174
  3. 2025FoundationStereo: Zero-Shot Stereo Matchingcited by 173
  4. 2022CroCo v2: Improved Cross-view Completion Pre-training for Stereo Matching and Optical Flowcited by 166
  5. 2021PENet: Towards Precise and Efficient Image Guided Depth Completioncited by 146
  6. 2018DeepLiDAR: Deep Surface Normal Guided Depth Prediction for Outdoor Scene from Sparse LiDAR Data and Single Color Imagecited by 129
  7. 2019Hierarchical Deep Stereo Matching on High-resolution Imagescited by 126
  8. 2019Learning Guided Convolutional Network for Depth Completioncited by 111
  9. 2018Depth Estimation via Affinity Learned with Convolutional Spatial Propagation Networkcited by 110
  10. 2024Selective-Stereo: Adaptive Frequency Information Selection for Stereo Matchingcited by 99
  11. 2018In Defense of Classical Image Processing: Fast Depth Completion on the CPUcited by 96
  12. 2018SegStereo: Exploiting Semantic Information for Disparity Estimationcited by 88
  13. 2017Learning Affinity via Spatial Propagation Networkscited by 79
  14. 2022Accurate and Efficient Stereo Matching via Attention Concatenation Volumecited by 59
  15. 2023LRRU: Long-short Range Recurrent Updating Networks for Depth Completioncited by 49
  16. 2023NeRF-Supervised Deep Stereocited by 46
  17. 2020FCFR-Net: Feature Fusion based Coarse-to-Fine Residual Learning for Depth Completioncited by 45
  18. 2022GraftNet: Towards Domain Generalized Stereo Matching with a Broad-Spectrum and Task-Oriented Featurecited by 43
  19. 2024Neural Markov Random Field for Stereo Matchingcited by 40
  20. 2019DFuseNet: Deep Fusion of RGB and Sparse Depth Information for Image Guided Dense Depth Completioncited by 38
  21. 2022Context-Enhanced Stereo Transformercited by 37
  22. 2023Uncertainty Guided Adaptive Warping for Robust and Efficient Stereo Matchingcited by 37
  23. 2019Noise-Aware Unsupervised Deep Lidar-Stereo Fusioncited by 31
  24. 2023OpenStereo: A Comprehensive Benchmark for Stereo Matching and Strong Baselinecited by 30
  25. 2019Guided Stereo Matchingcited by 24
  26. 2021A Surface Geometry Model for LiDAR Depth Completioncited by 18
  27. 20193D LiDAR and Stereo Fusion using Stereo Matching Network with Conditional Cost Volume Normalizationcited by 17
  28. 2021Volumetric Propagation Network: Stereo-LiDAR Fusion for Long-Range Depth Estimationcited by 15
  29. 2019LiStereo: Generate Dense Depth Maps from LIDAR and Stereo Imagerycited by 13
  30. 2024Federated Online Adaptation for Deep Stereocited by 12
  31. 2021S3: Learnable Sparse Signal Superdensity for Guided Depth Estimationcited by 10
  32. 2023Active Stereo Without Pattern Projectorcited by 8
  33. 2024Stereo-LiDAR Depth Estimation with Deformable Propagation and Learned Disparity-Depth Conversioncited by 4
  34. 2021Sparse LiDAR and Stereo Fusion (SLS-Fusion) for Depth Estimationand 3D Object Detectioncited by 3
  35. 2022Expansion of Visual Hints for Improved Generalization in Stereo Matchingcited by 3
  36. 2021Sparse LiDAR Assisted Self-supervised Stereo Disparity Estimationcited by 1

Built on by

Not cited yet.