Search papers, labs, and topics across Lattice.
This paper introduces Flose, a generative approach for 6D pose estimation that formulates the problem as conditional flow matching in $\mathbb{R}^3$. Flose overcomes limitations of regression-based and feature-matching methods by integrating appearance-based semantic features with geometric guidance for denoising, thereby handling object symmetries more effectively. The method incorporates RANSAC for outlier rejection and achieves a 4.5% improvement in Average Recall on the BOP benchmark compared to existing methods.
By blending geometric and appearance cues within a conditional flow matching framework, Flose cracks the long-standing problem of 6D pose estimation for symmetrical objects.
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/
