Quad-Networks: Unsupervised Learning to Rank for Interest Point Detection
Lists
Only you see this
Abstract
Several machine learning tasks require to represent the data using only a sparse set of interest points. An ideal detector is able to find the corresponding interest points even if the data undergo a transformation typical for a given domain. Since the task is of high practical interest in computer vision, many hand-crafted solutions were proposed. In this paper, we ask a fundamental question: can we learn such detectors from scratch? Since it is often unclear what points are interesting, human labelling cannot be used to find a truly unbiased solution. Therefore, the task requires an unsupervised formulation. We are the first to propose such a formulation: training a neural network to rank points in a transformation-invariant manner. Interest points are then extracted from the top/bottom quantiles of this ranking. We validate our approach on two tasks: standard RGB image interest point detection and challenging cross-modal interest point detection between RGB and depth images. We quantitatively show that our unsupervised method performs better or on-par with baselines.
BibTeX
@inproceedings{savinov2017quad,
author = {Nikolay Savinov and Akihito Seki and Lubor Ladicky and Torsten Sattler and Marc Pollefeys},
title = {{Quad-Networks: Unsupervised Learning to Rank for Interest Point Detection}},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition},
year = {2017},
doi = {10.1109/cvpr.2017.418},
}| Rank-SIFT: Learning to rank repeatable local interest points | CVPR · 2011 |
| SuperPoint: Self-Supervised Interest Point Detection and Description | Workshops @ CVPR · 2018 |
| SIPS: Unsupervised Succinct Interest Points | arXiv · 2018 |
| Neural Outlier Rejection for Self-Supervised Keypoint Learning | arXiv · 2019 |
| Learning Covariant Feature Detectors | arXiv · 2016 |
| The Maximal Self-dissimilarity Interest Point Detector | Ipsj Trans. Comput. Vis. Appl · 2015 |