Camera Distance-aware Top-down Approach for 3D Multi-person Pose\n Estimation from a Single RGB Image

2019/07/25 by Gyeongsik Moon, Moon, Gyeongsik, Ju Yong Chang +3 · 5 citations
Computer Science · #Anomaly Detection Techniques and Applications #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Human Pose and Action Recognition #Video Surveillance and Tracking Methods

paper · pdf · doi:10.48550/arxiv.1907.11346

openalex publication_date 2019/07/25 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Although significant improvement has been achieved recently in 3D human pose\nestimation, most of the previous methods only treat a single-person case. In\nthis work, we firstly propose a fully learning-based, camera distance-aware\ntop-down approach for 3D multi-person pose estimation from a single RGB image.\nThe pipeline of the proposed system consists of human detection, absolute 3D\nhuman root localization, and root-relative 3D single-person pose estimation\nmodules. Our system achieves comparable results with the state-of-the-art 3D\nsingle-person pose estimation models without any groundtruth information and\nsignificantly outperforms previous 3D multi-person pose estimation methods on\npublicly available datasets.\n The code is available in https://github.com/mks0601/3DMPPE_ROOTNET_RELEASE ,\nhttps://github.com/mks0601/3DMPPE_POSENET_RELEASE.\n

Cited by

• Proxy-Free Gaussian Splats Deformation with Splat-Based Surface Estimation
• Enhancing Multi-Camera Gymnast Tracking Through Domain Knowledge Integration
• AI-Enhanced Real-Time Wi-Fi Sensing Through Single Transceiver Pair
• Systematic Comparison of Projection Methods for Monocular 3D Human Pose Estimation on Fisheye Images
• PoseGRAF: Geometric-Reinforced Adaptive Fusion for Monocular 3D Human Pose Estimation

Related

• Multi-Person Absolute 3D Human Pose Estimation with Weak Depth Supervision
• Real-time marker-less multi-person 3D pose estimation in RGB-Depth camera networks
• SimpleDepthPose: Fast and Reliable Human Pose Estimation with RGBD-Images
• Multi-Person 3D Pose Estimation from Multi-View Uncalibrated Depth Cameras
• Single-Shot Multi-Person 3D Pose Estimation From Monocular RGB