GLAMpoints: Greedily Learned Accurate Match points

ICCV 2019

Prune Truong              Stefanos Apostolopoulos              Agata Mosinska           
Samuel Stucky                         Carlos Ciller                         Sandro De Zanet         

Abstract

We introduce a novel CNN-based feature point detector - GLAMpoints - learned in a semi-supervised manner. Our detector extracts repeatable, stable interest points with a dense coverage, specifically designed to maximize the correct matching in a specific domain, which is in contrast to conventional techniques that optimize indirect metrics. In this paper, we apply our method on challenging retinal slitlamp images, for which classical detectors yield unsatisfactory results due to low image quality and insufficient amount of low-level features. We show that GLAMpoints significantly outperforms classical detectors as well as state-of-the-art CNN-based methods in matching and registration quality for retinal images. Our method can also be extended to other domains, such as natural images. Training code and model weights are available here.

How to cite:

@inproceedings{Truong2019GLAMpoints,
          title={GLAMpoints: Greedily Learned Accurate Match Points},
          author={Prune Truong and Stefanos Apostolopoulos and Agata Mosinska and 
                  Samuel Stucky and Carlos Ciller and Sandro De Zanet},
          journal={2019 IEEE/CVF International Conference on Computer Vision (ICCV)},
          year={2019},
          pages={10731-10740}
}