mlsanityData to 14 Oct 2026

CA-I2P: Channel-Adaptive Registration Network with Global Optimal Selection

arXiv 2506.21364 ↗First posted 26 Jun 2025cs.CV cs.AIcited by 8

Not on the map. Why?

Abstract (arXiv)

Detection-free methods typically follow a coarse-to-fine pipeline, extracting image and point cloud features for patch-level matching and refining dense pixel-to-point correspondences. However, differences in feature channel attention between images and point clouds may lead to degraded matching results, ultimately impairing registration accuracy. Furthermore, similar structures in the scene could lead to redundant correspondences in cross-modal matching. To address these issues, we propose Channel Adaptive Adjustment Module (CAA) and Global Optimal Selection Module (GOS). CAA enhances intra-modal features and suppresses cross-modal sensitivity, while GOS replaces local selection with global optimization. Experiments on RGB-D Scenes V2 and 7-Scenes demonstrate the superiority of our method, achieving state-of-the-art performance in image-to-point cloud registration.

Key facts. Cited by 8 papers in the map (data through 14 Oct 2026). Builds on 41 of its 68 references that are in the map. First posted on arXiv 26 Jun 2025.

Builds on

41 of 68 references.

References · 41

  1. 2017Attention Is All You Needcited by 60,361
  2. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  3. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,248
  4. 2016Feature Pyramid Networks for Object Detectioncited by 5,036
  5. 2017Squeeze-and-Excitation Networkscited by 4,789
  6. 2016Improved Techniques for Training GANscited by 2,970
  7. 2020Unsupervised Learning of Visual Features by Contrasting Cluster Assignmentscited by 2,857
  8. 2018CBAM: Convolutional Block Attention Modulecited by 2,477
  9. 2013Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distancescited by 2,316
  10. 2019SuperGlue: Learning Feature Matching with Graph Neural Networkscited by 1,443
  11. 2019KPConv: Flexible and Deformable Convolution for Point Cloudscited by 1,205
  12. 2021Transformers in Vision: A Surveycited by 1,013
29 more
  1. 2021LoFTR: Detector-Free Local Feature Matching with Transformerscited by 984
  2. 2020Circle Loss: A Unified Perspective of Pair Similarity Optimizationcited by 403
  3. 2020PREDATOR: Registration of 3D Point Clouds with Low Overlapcited by 290
  4. 2017Joint Distribution Optimal Transportation for Domain Adaptationcited by 282
  5. 2017Hand Keypoint Detection in Single Images using Multiview Bootstrappingcited by 231
  6. 2022Geometric Transformer for Fast and Robust Point Cloud Registrationcited by 202
  7. 2021OTA: Optimal Transport Assignment for Object Detectioncited by 178
  8. 2021CoFiNet: Reliable Coarse-to-fine Correspondences for Robust Point Cloud Registrationcited by 152
  9. 2024Efficient LoFTR: Semi-Dense Local Feature Matching with Sparse-Like Speedcited by 120
  10. 2019Key.Net: Keypoint Detection by Handcrafted and Learned CNN Filterscited by 118
  11. 2023GeoTransformer: Fast and Robust Point Cloud Registration with Geometric Transformercited by 84
  12. 2020Rotate to Attend: Convolutional Triplet Attention Modulecited by 75
  13. 20192D3D-MatchNet: Learning to Match Keypoints Across 2D Image and 3D Point Cloudcited by 62
  14. 2022Revisiting Domain Generalized Stereo Matching Networks from a Feature Consistency Perspectivecited by 46
  15. 2023Learning Cut Selection for Mixed-Integer Linear Programming via Hierarchical Sequence Modelcited by 45
  16. 2022CorrI2P: Deep Image-to-Point Cloud Registration via Dense Correspondencecited by 44
  17. 2021P2-Net: Joint Description and Detection of Local Features for Pixel and Point Matchingcited by 42
  18. 2023FreeReg: Image-to-Point Cloud Registration Leveraging Pretrained Diffusion Models and Monocular Depth Estimatorscited by 31
  19. 20232D3D-MATR: 2D-3D Matching Transformer for Detection-free Registration between Images and Point Cloudscited by 27
  20. 2023SE-ORNet: Self-Ensembling Orientation-aware Network for Unsupervised Point Cloud Shape Correspondencecited by 19
  21. 2025Bridge 2D-3D: Uncertainty-aware Hierarchical Registration Network with Domain Alignmentcited by 14
  22. 2025Brightness Perceiving for Recursive Low-Light Image Enhancementcited by 14
  23. 2014Generalized Max Poolingcited by 12
  24. 2024Learning to Cut via Hierarchical Sequence/Set Model for Efficient Mixed-Integer Programmingcited by 12
  25. 2024Learning to Stop Cut Generation for Efficient Mixed-Integer Linear Programmingcited by 8
  26. 2024Unsupervised Template-assisted Point Cloud Shape Correspondence Networkcited by 7
  27. 2021Comparison of Methods Generalizing Max- and Average-Poolingcited by 6
  28. 2025State Space Model Meets Transformer: A New Paradigm for 3D Object Detectioncited by 5
  29. 2024Diff-Reg v1: Diffusion Matching Model for Registration Problemcited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 8.

8 papers.

  1. 2026VCR: Variance-Driven Channel Recalibration for Robust Low-Light Enhancementcited by 1
  2. 2026GeoGuide: Hierarchical Geometric Guidance for Open-Vocabulary 3D Semantic Segmentationcited by 1
  3. 2026GLASS: Geometry-aware Local Alignment and Structure Synchronization Network for 2D-3D Registrationcited by 1
  4. 2026ReFlow: Self-correction Motion Learning for Dynamic Scene Reconstructioncited by 0
  5. 2026FS-I2P:A Hierarchical Focus-Sweep Registration Network with Dynamically Allocated Depthcited by 0
  6. 2026Cross-Coordinate Correspondence Pruning for Image-to-Point Cloud Registrationcited by 0
  7. 2026Mask 2D-3D: Adaptive Dual-Masked Autoencoder Network for Image-to-Point Cloud Registrationcited by 0
  8. 2026OCA: ODE-Driven Cross-Attention for Image-to-Point-Cloud Registrationcited by 0