FPC-Net: Revisiting SuperPoint with Descriptor-Free Keypoint Detection via Feature Pyramids and Consistency-Based Implicit Matching

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Grigore, Ionuţ, Popa, Călin-Adrian, Leoveanu-Condrei, Claudiu
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908449530445824
author Grigore, Ionuţ
Popa, Călin-Adrian
Leoveanu-Condrei, Claudiu
author_facet Grigore, Ionuţ
Popa, Călin-Adrian
Leoveanu-Condrei, Claudiu
contents The extraction and matching of interest points are fundamental to many geometric computer vision tasks. Traditionally, matching is performed by assigning descriptors to interest points and identifying correspondences based on descriptor similarity. This work introduces a technique where interest points are inherently associated during detection, eliminating the need for computing, storing, transmitting, or matching descriptors. Although the matching accuracy is marginally lower than that of conventional approaches, our method completely eliminates the need for descriptors, leading to a drastic reduction in memory usage for localization systems. We assess its effectiveness by comparing it against both classical handcrafted methods and modern learned approaches.
format Preprint
id arxiv_https___arxiv_org_abs_2507_10770
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FPC-Net: Revisiting SuperPoint with Descriptor-Free Keypoint Detection via Feature Pyramids and Consistency-Based Implicit Matching
Grigore, Ionuţ
Popa, Călin-Adrian
Leoveanu-Condrei, Claudiu
Computer Vision and Pattern Recognition
The extraction and matching of interest points are fundamental to many geometric computer vision tasks. Traditionally, matching is performed by assigning descriptors to interest points and identifying correspondences based on descriptor similarity. This work introduces a technique where interest points are inherently associated during detection, eliminating the need for computing, storing, transmitting, or matching descriptors. Although the matching accuracy is marginally lower than that of conventional approaches, our method completely eliminates the need for descriptors, leading to a drastic reduction in memory usage for localization systems. We assess its effectiveness by comparing it against both classical handcrafted methods and modern learned approaches.
title FPC-Net: Revisiting SuperPoint with Descriptor-Free Keypoint Detection via Feature Pyramids and Consistency-Based Implicit Matching
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2507.10770