室内环境下无校准的3D多摄像机人体跟踪
Calibration-Free 3D Multi-Camera People Tracking for Indoor Environment
浏览论文内容
中文总结 AI 辅助
针对室内环境下多摄像机人体跟踪依赖校准的问题，提出统一无校准3D MCPT框架，集成多种技术，采用姿态引导3D提升策略和全局身份关联方法，在挑战赛中取得53.13%的HOTA分数，为纯视觉3D跟踪建立基线。
中文摘要 AI 辅助
多摄像机人体跟踪（MCPT）传统上依赖精确的相机内参和外参校准，将2D检测投影到统一的3D世界坐标中。手动校准是从无约束视频档案生成大规模数据集的主要瓶颈。本文提出了一个统一的无校准3D MCPT框架，该框架使用深度基础模型直接从视觉数据推断几何结构。系统集成了无锚点检测（YOLOX）、鲁棒跟踪（BoT-SORT）、全尺度外观嵌入（OsNet）、姿态估计（通过MMPose的HRNet）以及使用视觉几何基础变换器（VGGT）的基于变换器的几何重建。一种姿态引导的3D提升策略将头部关键点投影到重建流形上，消除了对地面平面单应性的依赖。全局身份关联被制定为在具有严格速度门控的联合外观-几何成本下的层次凝聚聚类。在2024年人工智能城市挑战赛上的评估表明，在无法获取地面真值校准矩阵的情况下，HOTA分数为53.13%，为基于纯视觉的3D跟踪建立了一个强大的基线。
英文摘要
Multi-Camera People Tracking (MCPT) traditionally relies on precise intrinsic and extrinsic camera calibration to project 2D detections into a unified 3D world coordinate system.However, manual calibration constitutes a major bottleneck in large-scale dataset generation from unconstrained video archives. This work proposes a unified calibration-free 3D MCPT framework that infers geometric structure directly from visual data using deep foundation models. The system integrates anchor-free detection (YOLOX), robust tracking (BoT-SORT), omni-scale appearance embedding (OsNet), pose estimation (HRNet via MMPose), and transformer-based geometric reconstruction using the Visual Geometry Grounded Transformer (VGGT). A pose-guided 3D lifting strategy projects head keypoints onto a reconstructed manifold, eliminating dependence on ground-plane homography. Global identity association is formulated as hierarchical agglomerative clustering under a joint appearance-geometry cost with strict velocity gating. Evaluation on the AI City Challenge 2024 demonstrates a HOTA score of 53.13% without access to ground-truth calibration matrices, establishing a strong baseline for purely vision-based 3D tracking.
