.png){kind=icon}
.png){kind=icon}
Return
MHDPose: Multi-hypothesis 3D human pose estimation using bidirectional Mamba diffusion models
.png){kind=icon}
OA
AI
DOI：10.1016/j.patcog.2026.114396.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• We propose a diffusion-based multi-hypothesis framework for 3D human pose estimation. • We designed a hybrid denoiser architecture combining transformer and state space model. • We introduce kinematic-aware spatial encoding to improve joint reasoning. • We design a probabilistic approach for generating diverse 3D poses from 2D observations. • Experimental results show improved accuracy on benchmark datasets.
Keywords:
Human pose estimation
Diffusion
Mamba
State Space Models
Multi-hypothesis generation
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.