Generative 6D Pose Estimation via Conditional Flow Matching

TL;DR

Flose introduces a novel generative approach for 6D pose estimation that leverages conditional flow matching with appearance-based features, outperforming prior methods on multiple datasets.

Contribution

The paper presents Flose, a new generative method using conditional flow matching with semantic features for robust 6D pose estimation, especially in symmetric objects.

Findings

Flose achieves +4.5 average recall over prior methods.

Incorporates appearance features to resolve symmetries.

Validated on five BOP benchmark datasets.

Abstract

Existing methods for instance-level 6D pose estimation typically rely on neural networks that either directly regress the pose in $\mathrm{SE}(3)$ or estimate it indirectly via local feature matching. The former struggle with object symmetries, while the latter fail in the absence of distinctive local features. To overcome these limitations, we propose a novel formulation of 6D pose estimation as a conditional flow matching problem in $\mathbb{R}^3$. We introduce Flose, a generative method that infers object poses via a denoising process conditioned on local features. While prior approaches based on conditional flow matching perform denoising solely based on geometric guidance, Flose integrates appearance-based semantic features to mitigate ambiguities caused by object symmetries. We further incorporate RANSAC-based registration to handle outliers. We validate Flose on five datasets from the established BOP benchmark. Flose outperforms prior methods with an average improvement of +4.5 Average Recall. Project Website : https://tev-fbk.github.io/Flose/