PulseAugur
LIVE 07:05:28
research · [1 source] ·
0
research

Geometry-conditioned diffusion model enhances occlusion-robust pose estimation

Researchers have developed a new method called Geometry-Conditioned Diffusion (GeoDiffPose) to improve human pose estimation in bed, particularly when parts of the body are covered by blankets. This approach uses a pose-conditioned Latent Diffusion Model to generate images of covered poses directly from skeletal keypoints, bypassing the need for paired supervision or visible source imagery. The GeoDiffPose model demonstrates effectiveness in occlusion-robust pose estimation, achieving high localization accuracy under severe occlusion and approaching the performance of fully supervised methods. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces a novel diffusion model approach for occlusion-robust pose estimation, potentially improving applications in healthcare and surveillance.

RANK_REASON This is a research paper detailing a new method for pose estimation.

Read on arXiv cs.CV →

COVERAGE [1]

  1. arXiv cs.CV TIER_1 · Navid Aslankhani Khameneh, Marco Carletti, Cigdem Beyan ·

    Geometry-Conditioned Diffusion for Occlusion-Robust In-Bed Pose Estimation

    arXiv:2604.23651v1 Announce Type: new Abstract: Robust in-bed human pose estimation under blanket occlusion remains challenging due to the scarcity of reliable labeled training data for heavily covered poses. Existing approaches rely on multi-modal sensing or image-to-image trans…