AI Paper Atlas

Research map / 3D vision / Human pose estimation

Occluded Human Body Capture with Frequency Domain Denoising Prior

Buzhen Huang, Chongyang Xu, Wentao Tang, Yuan Shu, Jingyi Ju, Binghui Zuo, Yangang Wang

CVPR 2026 · 3D vision › Human pose estimation · approach: Physics and sensor-based motion capture

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add pretrained frequency-domain denoising prior encoder to the 3D lifting branch (ablation)Temporal Regression · OcMotionMPJPE: 51.7 -> 49.1Improved
train 3D decoder with the same diffusion process via reprojected keypoints (ablation)Prior model · OcMotionMPJPE: 49.1 -> 48.5Improved
use DWT instead of DCT for denoising keypoints (ablation)Denoised Keypoints · OcMotionMPJPE: 51.6 -> 49.6Improved
use DCT instead of DWT for denoising keypoints (ablation)Denoised Keypoints · OcMotionAccel.: 18.8 -> 17.6Improved
add detected 2D keypoints from ViTPose (ablation)Temporal Regression · OcMotionMPJPE: 51.7 -> 52.3Hurt
add time-domain keypoint denoising with temporal transformer (ablation)Temporal Regression · OcMotionAccel.: 28.2 -> 23.3Improved
add DWT-based denoising of keypoints instead of time-domain denoising (ablation)Denoised Keypoints · OcMotionMPJPE: 51.4 -> 49.6Improved
add plain 3D diffusion lifting (ablation)Temporal Regression + predicted keypoints · OcMotionMPJPE: 52.3 -> 51.3No clear effect
apply DWT to 3D diffusion without pretrained 2D prior (ablation)3D Diffusion · OcMotionMPJPE: 51.3 -> 51.0No clear effect
add ground-truth 2D keypoints (oracle) to temporal regression (ablation)Temporal Regression · OcMotionMPJPE: 51.7 -> 32.5Improved
replace frame-wise regression by temporal transformer regression (ablation)Frame Regression · OcMotionAccel.: 31.4 -> 28.2Improved
use frequency-domain method instead of CloseInt multi-person interaction methodHi4D inter-person occlusion · Hi4DMPJPE: 63.1 -> 61.5Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy