{
  "paper_id": "6c00db1e2b31ee06b71d1b8404f8fbe8",
  "paper_title": "MARKERLESS MULTI-VIEW 3D HUMAN POSE ESTIMATION: A SURVEY",
  "paper_authors": [
    "Ana Filipa Rodrigues Nogueira",
    "Hélder P. Oliveira",
    "Luís F. Teixeira"
  ],
  "publication_date": "2024",
  "publication_venue": "Image and Vision Computing",
  "abstract": "3D human pose estimation involves reconstructing the human skeleton by detecting the body joints. Accurate and efficient solutions are required for several real-world applications including animation, human-robot interaction, surveillance, and sports. However, challenges such as occlusions, 2D pose mismatches, random camera perspectives, and limited 3D labelled data have been hampering the models' performance and limiting their deployment in real-world scenarios. The higher availability of cameras has led researchers to explore multi-view solutions to take advantage of the different perspectives to reconstruct the pose. Most existing reviews have mainly focused on monocular 3D human pose estimation, so a comprehensive survey on multi-view approaches has been missing since 2012. According to the reviewed articles, the majority of the existing methods are fully-supervised approaches based on geometric constraints, which are often limited by 2D pose mismatches. To mitigate this, researchers have proposed incorporating temporal consistency or depth information. Alternatively, working directly with 3D features has been shown to completely overcome this issue, albeit at the cost of increased computational complexity. Additionally, models with lower levels of supervision have been identified to help address challenges such as annotated data scarcity and generalisation to new setups. Therefore, no method currently addresses all challenges associated with 3D pose reconstruction, and a trade-off between complexity and performance exists. Further research is needed to develop approaches capable of quickly inferring a highly accurate 3D pose with bearable computation cost. Techniques such as active learning, low-supervision methods, temporal consistency, view selection, depth information estimation, and multi-modal approaches are strategies to keep in mind when developing a new methodology to solve this task.",
  "keywords": [
    "3D Human Pose Estimation",
    "Multi-view",
    "Supervision level",
    "Temporal consistency",
    "Multi-modal"
  ],
  "paper_institutions": [
    "Instituto de Engenharia de Sistemas e Computadores, Tecnologia e Ciência (INESC TEC)",
    "Faculdade de Engenharia da Universidade do Porto (FEUP)",
    "Faculdade de Ciências da Universidade do Porto (FCUP)"
  ],
  "output_language": "zh",
  "provider": "doubao",
  "model": "ark-code-latest",
  "prompt_template": "simple",
  "extracted_at": "2026-08-16T19:47:12.067540Z",
  "template_tag": "simple",
  "summary": "### 摘要\n本文系统综述了基于多视图的无标记三维人体姿态估计方法。该任务旨在从多个摄像机的图像中重建人体骨架，应用于动画、人机交互、监控、体育等多个领域。现有方法面临遮挡、二维姿态估计误差、相机视角不固定以及三维标注数据稀缺等挑战。大多数方法采用几何约束的监督学习，先估计各视角二维姿态，再通过三角化等方法重建三维姿态，但二维姿态误差和跨视角匹配问题限制了性能。为缓解这些问题，研究者引入时间一致性、深度信息，或直接使用三维特征（如体素方法），后者完全避免二维匹配错误但计算复杂度高。此外，低监督方法（半监督、自监督等）有助于缓解标注数据稀缺和泛化问题。本文回顾了2012年以来的相关文献，涵盖单人和多人场景，并将方法分类为全监督、半监督至无监督，以及几何约束、体素和平面扫描立体三类。评测常用Human3.6M、Campus、Shelf等数据集，指标包括MPJPE、PCP、PCK等。结果发现，无单一方法能同时兼顾高精度、快速推理和低计算成本，存在性能与复杂度的权衡。未来研究方向包括主动学习、低监督方法、时间一致性、视角选择、深度估计和多模态融合。\n\n### 关键要点\n- 无标记多视角三维人体姿态估计是计算机视觉的重要方向，应用广泛，但存在遮挡、二维姿态误差和标注数据不足等挑战。\n- 经典流程是逐视角估计二维姿态，然后通过几何关联（如三角化）构建三维骨架，但该方法严重依赖相机标定和视角数量。\n- 全监督方法在Human3.6M上最佳MPJPE可达15.1mm（FusionFormer），但需大量标注数据，泛化能力有限。\n- 基于体素的方法（如VoxelPose）直接在三维空间操作，避免了二维匹配问题，但计算开销大，推理速度慢。\n- 平面扫描立体方法（如Lin和Lee提出的）在二维姿态基础上引入深度信息，在精度和速度间取得平衡，在Campus和Shelf数据集上PCP优于传统体素方法。\n- 低监督方法（半监督、自监督）可在缺乏三维标注时工作，例如Liu等人的自监督方法在Human3.6M上MPJPE达22.5mm，接近全监督最佳结果。\n- 时间一致性被广泛用于多人跟踪和减少误检，但可能抑制快速微小动作，在某些应用如监控中需谨慎。\n- 缓解遮挡可选用视点选择策略，或使用RGB-D、ToF、360°相机等提供深度信息的传感器。\n- 无线信号（如Wi-Fi、RF）可避免光照和遮挡问题，但现有工作仅将相机用于生成真值，未融合多模态信息。\n- 主动学习可降低标注成本，但该策略在文献中较少应用，其潜力尚待发掘。\n- 综述发现，没有方法能在高精度、快速推理和低计算成本方面同时最优，实际应用需根据场景权衡。"
}