SRPose: Two-view Relative Pose Estimation with Sparse Keypoints

TL;DR

SRPose is a novel sparse keypoint-based framework for two-view relative pose estimation that is accurate, fast, and adaptable to various scenarios and camera settings, outperforming existing methods.

Contribution

The paper introduces SRPose, a new framework combining sparse keypoints, position encoding, and promptable attention for versatile and efficient relative pose estimation.

Findings

Achieves state-of-the-art accuracy and speed.

Generalizes well to different scenarios and camera parameters.

Robust to varying image sizes and low-resource environments.

Abstract

Two-view pose estimation is essential for map-free visual relocalization and object pose tracking tasks. However, traditional matching methods suffer from time-consuming robust estimators, while deep learning-based pose regressors only cater to camera-to-world pose estimation, lacking generalizability to different image sizes and camera intrinsics. In this paper, we propose SRPose, a sparse keypoint-based framework for two-view relative pose estimation in camera-to-world and object-to-camera scenarios. SRPose consists of a sparse keypoint detector, an intrinsic-calibration position encoder, and promptable prior knowledge-guided attention layers. Given two RGB images of a fixed scene or a moving object, SRPose estimates the relative camera or 6D object pose transformation. Extensive experiments demonstrate that SRPose achieves competitive or superior performance compared to state-of-the-art methods in terms of accuracy and speed, showing generalizability to both scenarios. It is robust to different image sizes and camera intrinsics, and can be deployed with low computing resources.