Learnable Motion Coherence for Correspondence Pruning

  title={Learnable Motion Coherence for Correspondence Pruning},
  author={Yuan Liu and Lingjie Liu and Chu-Hsing Lin and Zhen Dong and Wenping Wang},
  journal={2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  • Yuan Liu, Lingjie Liu, +2 authors Wenping Wang
  • Published 30 November 2020
  • Computer Science
  • 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Motion coherence is an important clue for distinguishing true correspondences from false ones. Modeling motion coherence on sparse putative correspondences is challenging due to their sparsity and uneven distributions. Existing works on motion coherence are sensitive to parameter settings and have difficulty in dealing with complex motion patterns. In this paper, we introduce a network called Laplacian Motion Coherence Network (LMCNet) to learn motion coherence property for correspondence… Expand
SPATL: Salient Parameter Aggregation and Transfer Learning for Heterogeneous Clients in Federated Learning
This paper introduces a salient parameter selection agent based on deep reinforcement learning on local clients, and aggregating the selected salient parameters on the central server, and leverages the gradient control mechanism to correct the gradient heterogeneity among clients. Expand


Learning to Find Good Correspondences
A novel normalization technique, called Context Normalization, is introduced, which allows the network to process each data point separately while embedding global information in it, and also makes the network invariant to the order of the correspondences. Expand
Learning Two-View Correspondences and Geometry Using Order-Aware Network
This paper proposes Order-Aware Network, which infers the probabilities of correspondences being inliers and regresses the relative pose encoded by the essential matrix, and is built hierarchically and comprises three novel operations. Expand
CODE: Coherence Based Decision Boundaries for Feature Correspondence
  • Wen-Yan Lin, Fan Wang, +4 authors Jiangbo Lu
  • Computer Science, Medicine
  • IEEE Transactions on Pattern Analysis and Machine Intelligence
  • 2018
A non-linear regression technique is proposed that can discover a coherence based separability constraint from highly noisy matches and embed it into a correspondence likelihood model and integrate the model into a full feature correspondence system which reliably generates large numbers of good quality correspondences over wide baselines. Expand
Neighbourhood Consensus Networks
An end-to-end trainable convolutional neural network architecture that identifies sets of spatially consistent matches by analyzing neighbourhood consensus patterns in the 4D space of all possible correspondences between a pair of images without the need for a global geometric model is developed. Expand
Modeling temporal coherence for optical flow
A novel parametrization for multi-frame optical flow computation that naturally enables us to embed the assumption of a temporally coherent spatial flow structure, as well as the assumption that the optical flow is smooth along motion trajectories. Expand
The Robust Estimation of Multiple Motions: Parametric and Piecewise-Smooth Flow Fields
A framework based on robust estimation is presented that addresses violations of the brightness constancy and spatial smoothness assumptions caused by multiple motions of optical flow, and is applied to standard formulations of the optical flow problem thus reducing their sensitivity to violations of their underlying assumptions. Expand
Non-rigid point set registration: Coherent Point Drift
This work formulate the motion coherence constraint and derive a solution of regularized ML estimation through the variational approach, which leads to an elegant kernel form, and derives the EM algorithm for the penalized ML optimization with deterministic annealing. Expand
Learning to Guide Local Feature Matches
A learning-based approach to guide local feature matches via a learned approximate image matching can boost the results of SIFT to a level similar to state-of-the-art deep descriptors, such as Superpoint, ContextDesc, or D2-Net and can improve performance for these descriptors. Expand
SuperGlue: Learning Feature Matching With Graph Neural Networks
SuperGlue is introduced, a neural network that matches two sets of local features by jointly finding correspondences and rejecting non-matchable points and introduces a flexible context aggregation mechanism based on attention, enabling SuperGlue to reason about the underlying 3D scene and feature assignments jointly. Expand
Efficient Neighbourhood Consensus Networks via Submanifold Sparse Convolutions
The proposed Sparse-NCNet method obtains state-of-the-art results on the HPatches Sequences and InLoc visual localisation benchmarks, and competitive results in the Aachen Day-Night benchmark. Expand