PulseAugur
EN
LIVE 12:01:51

DreamSat-Pose framework estimates spacecraft pose from single images

Researchers have developed DreamSat-Pose, a new framework for estimating the six-degrees-of-freedom pose of unknown spacecraft from a single image. The system first reconstructs a 3D shape model of the target and then uses learned 2D-3D correspondences to determine its pose. Utilizing a DINOv3 vision transformer for image features and a graph convolutional neural network for geometric features, the method achieves a mean pointing error of 0.157 degrees on the SPE3R dataset, outperforming the FoundationPose baseline. AI

IMPACT This research advances single-view 3D reconstruction and pose estimation techniques, potentially improving autonomous operations in space.

RANK_REASON The cluster contains a research paper detailing a new method for pose estimation.

Read on arXiv cs.LG →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

DreamSat-Pose framework estimates spacecraft pose from single images

COVERAGE [2]

  1. arXiv cs.LG TIER_1 English(EN) · Josiane Uwumukiza, Jocelyn Zhao, Giovanni Lavezzi, Giacomo Battaglia, Paolo Panicucci, Minduli C. Wijayatunga, Victor Rodriguez-Fernandez, Richard Linares ·

    DreamSat-Pose: Spacecraft Pose Estimation from Single-View 3D Reconstructions and Learned 2D-3D Feature Matching

    arXiv:2607.13449v1 Announce Type: cross Abstract: 6-DoF pose estimation is a critical task in autonomous rendezvous and proximity operations. In the case of an unknown target, this task becomes challenging as it shall be paired with the reconstruction of the target shape model. I…

  2. arXiv cs.LG TIER_1 English(EN) · Richard Linares ·

    DreamSat-Pose: Spacecraft Pose Estimation from Single-View 3D Reconstructions and Learned 2D-3D Feature Matching

    6-DoF pose estimation is a critical task in autonomous rendezvous and proximity operations. In the case of an unknown target, this task becomes challenging as it shall be paired with the reconstruction of the target shape model. In this article, we propose a novel framework for s…