QuadricSLAM: Dual Quadrics from Object Detections as Landmarks in Object-oriented SLAM

TL;DR

This paper introduces a SLAM approach that uses 2D object detections to estimate 3D quadrics as landmarks, enabling accurate object representation and camera localization in a unified framework.

Contribution

It proposes a novel SLAM formulation using dual quadrics as landmarks derived from 2D object detections, including a geometric error formulation and a sensor model for partial visibility.

Findings

Effective joint estimation of camera pose and object quadrics.

Robust handling of partially visible objects in SLAM.

Demonstrated improved object and camera localization accuracy.

Abstract

In this paper, we use 2D object detections from multiple views to simultaneously estimate a 3D quadric surface for each object and localize the camera position. We derive a SLAM formulation that uses dual quadrics as 3D landmark representations, exploiting their ability to compactly represent the size, position and orientation of an object, and show how 2D object detections can directly constrain the quadric parameters via a novel geometric error formulation. We develop a sensor model for object detectors that addresses the challenge of partially visible objects, and demonstrate how to jointly estimate the camera pose and constrained dual quadric parameters in factor graph based SLAM with a general perspective camera.