Correspondence Networks with Adaptive Neighbourhood Consensus
1Active Vision Lab & 2Visual Geometry Group
Department of Engineering Science, University of Oxford
* indicates equal contribution
|
Abstract
In this paper, we tackle the task of establishing dense visual correspondences between images containing
objects of the same category.
This is a challenging task due to large intra-class variations and a lack of dense pixel level annotations.
We propose a convolutional neural network architecture, called adaptive neighbourhood consensus network
(ANC-Net), that can be trained end-to-end with sparse key-point annotations, to handle this challenge.
At the core of ANC-Net is our proposed non-isotropic 4D convolution kernel, which forms the building block
for the adaptive neighbourhood consensus module for robust matching.
We also introduce a simple and efficient multi-scale self-similarity module in ANC-Net to make the learned
feature robust to intra-class variations.
Furthermore, we propose a novel orthogonal loss that can enforce the one-to-one matching constraint.
We thoroughly evaluate the effectiveness of our method on various benchmarks, where it substantially
outperforms state-of-the-art methods.
BibTex
@inproceedings{Li2020Correspondence,
author = {Shuda Li and Kai Han and Theo W. Costain and Henry Howard-Jenkins and Victor Prisacariu},
title = {Correspondence Networks with Adaptive Neighbourhood Consensus},
booktitle = {IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
year = {2020},
}
Acknowledgments
We gratefully acknowledge the support of the European Commission Project Multiple-actOrs Virtual
EmpathicCARegiver for the Elder (MoveCare) and the EPSRC Programme Grant Seebibyte EP/M013774/1.
|
Webpage template borrowed from Split-Brain Autoencoders,
CVPR 2017.