DISK: Learning local features with policy gradient

TL;DR

DISK introduces a reinforcement learning-based approach to learn local features end-to-end, enabling dense, discriminative keypoints that achieve state-of-the-art results on benchmark datasets.

Contribution

The paper presents DISK, a novel RL-based method for end-to-end learning of local features, overcoming traditional discrete selection challenges.

Findings

Achieves state-of-the-art results on three benchmarks.

Enables dense, discriminative keypoint extraction.

Challenges assumptions about keypoint quality.

Abstract

Local feature frameworks are difficult to learn in an end-to-end fashion, due to the discreteness inherent to the selection and matching of sparse keypoints. We introduce DISK (DIScrete Keypoints), a novel method that overcomes these obstacles by leveraging principles from Reinforcement Learning (RL), optimizing end-to-end for a high number of correct feature matches. Our simple yet expressive probabilistic model lets us keep the training and inference regimes close, while maintaining good enough convergence properties to reliably train from scratch. Our features can be extracted very densely while remaining discriminative, challenging commonly held assumptions about what constitutes a good keypoint, as showcased in Fig. 1, and deliver state-of-the-art results on three public benchmarks.