Weakly supervised learning for image keypoint matching using graph convolutional networks

作者：

Highlights：

• We propose a novel image matching system by leveraging feature knowledge.

• We consider three key factors to improve performance of image feature matching.

• Graph convolutional networks directly use image keypoints as input data.

• Multi-level feature knowledge representations can discriminate different keypoints.

摘要

•We propose a novel image matching system by leveraging feature knowledge.•We consider three key factors to improve performance of image feature matching.•Graph convolutional networks directly use image keypoints as input data.•Multi-level feature knowledge representations can discriminate different keypoints.

论文关键词：Feature matching,Keypoints,Mismatch removal,Deep neural networks

论文评审过程：Received 8 August 2019, Revised 2 April 2020, Accepted 2 April 2020, Available online 11 April 2020, Version of Record 24 April 2020.

论文官网地址：https://doi.org/10.1016/j.knosys.2020.105871