XFeat: Accelerated Features for Lightweight Image Matching

TL;DR

XFeat is a lightweight, fast, and versatile neural network architecture for local image feature detection and matching, suitable for resource-limited devices and supporting semi-dense matching with high accuracy.

Contribution

It introduces a novel, efficient architecture for local feature matching that supports semi-dense matching and runs in real-time on CPU without hardware optimization.

Findings

XFeat is up to 5x faster than existing methods.

It achieves comparable or better accuracy in pose estimation and localization.

It operates in real-time on inexpensive CPU hardware.

Abstract

We introduce a lightweight and accurate architecture for resource-efficient visual correspondence. Our method, dubbed XFeat (Accelerated Features), revisits fundamental design choices in convolutional neural networks for detecting, extracting, and matching local features. Our new model satisfies a critical need for fast and robust algorithms suitable to resource-limited devices. In particular, accurate image matching requires sufficiently large image resolutions - for this reason, we keep the resolution as large as possible while limiting the number of channels in the network. Besides, our model is designed to offer the choice of matching at the sparse or semi-dense levels, each of which may be more suitable for different downstream applications, such as visual navigation and augmented reality. Our model is the first to offer semi-dense matching efficiently, leveraging a novel match refinement module that relies on coarse local descriptors. XFeat is versatile and hardware-independent, surpassing current deep learning-based local features in speed (up to 5x faster) with comparable or better accuracy, proven in pose estimation and visual localization. We showcase it running in real-time on an inexpensive laptop CPU without specialized hardware optimizations. Code and weights are available at www.verlab.dcc.ufmg.br/descriptors/xfeat_cvpr24.