Correspondence Estimation

Correspondence estimation, the task of identifying matching points between different views or representations of a scene, is crucial for numerous computer vision and robotics applications, such as 3D reconstruction, visual odometry, and object pose estimation. Current research focuses on improving the accuracy and robustness of correspondence estimation, particularly in challenging scenarios like incomplete data, cross-domain variations, and occlusions, employing techniques like recurrent neural networks, differentiable bundle adjustment, and multi-scale feature fusion within various model architectures. These advancements are driving progress in areas such as autonomous driving, augmented reality, and industrial automation by enabling more accurate and reliable perception and scene understanding.

Papers

July 6, 2024

T-CorresNet: Template Guided 3D Point Cloud Completion with Correspondence Pooling Query Generation Strategy
Fan Duan, Jiahao Yu, Li Chen
Point Cloud Point Cloud Completion Deep Template Spherical Gaussian Correspondence Estimation

June 12, 2024

From Variance to Veracity: Unbundling and Mitigating Gradient Variance in Differentiable Bundle Adjustment Layers
Swaminathan Gurumurthy, Karnik Ram, Bingqing Chen, Zachary Manchester, Zico Kolter
Visual Odometry Bundle Adjustment Variance Information Pose Estimator Bundle Sale Data High Quality Correspondence Correspondence Estimation

January 16, 2024

Cross-Domain Few-Shot Segmentation via Iterative Support-Query Correspondence Mining
Jiahao Nie, Yun Xing, Gongjie Zhang, Pei Yan, Aoran Xiao, Yap-Peng Tan, Alex C. Kot, Shijian Lu
Fine Tuning Segmentation Based Approach Cross Domain Shot Adaptation Correspondence Estimation

December 6, 2022

Self-Supervised Correspondence Estimation via Multiview Registration
Mohamed El Banani, Ignacio Rocco, David Novotny, Andrea Vedaldi, Natalia Neverova, Justin Johnson, Benjamin Graham
Correspondence Learning Consecutive Frame RGB D Sequence Self Supervised Correspondence Correspondence Estimation Multiview Registration

August 31, 2022

Improving RGB-D Point Cloud Registration by Learning Multi-scale Local Linear Transformation
Ziming Wang, Xiaoliang Huo, Zhenghao Chen, Jing Zhang, Lu Sheng, Dong Xu
Point Cloud Registration Correspondence Estimation

August 21, 2022

SIM2E: Benchmarking the Group Equivariant Capability of Correspondence Matching Algorithms
Shuai Su, Zhongkai Zhao, Yixin Fei, Shuda Li, Qijun Chen, Rui Fan
Equivariant Map Based Correspondence Correspondence Matching Robust Correspondence Correspondence Estimation

March 24, 2022

RNNPose: Recurrent 6-DoF Object Pose Refinement with Robust Correspondence Field Estimation and Pose Optimization
Yan Xu, Kwan-Yee Lin, Guofeng Zhang, Xiaogang Wang, Hongsheng Li
Recurrent Event Object Pose Pose Refinement Pose Optimization Based Correspondence Correspondence Estimation

Correspondence Estimation

Papers

T-CorresNet: Template Guided 3D Point Cloud Completion with Correspondence Pooling Query Generation Strategy

From Variance to Veracity: Unbundling and Mitigating Gradient Variance in Differentiable Bundle Adjustment Layers

Cross-Domain Few-Shot Segmentation via Iterative Support-Query Correspondence Mining

Self-Supervised Correspondence Estimation via Multiview Registration

Improving RGB-D Point Cloud Registration by Learning Multi-scale Local Linear Transformation

SIM2E: Benchmarking the Group Equivariant Capability of Correspondence Matching Algorithms

RNNPose: Recurrent 6-DoF Object Pose Refinement with Robust Correspondence Field Estimation and Pose Optimization