BigMaQ (Big MacaQue 3D Motion and Animation Dataset)
收藏资源简介:
BigMaQ是由蒂宾根大学等机构联合构建的首个结合3D表面建模的猕猴动作数据集,包含超过750个互动场景的16视角同步视频与骨骼关节旋转数据。该数据集通过16台校准相机采集,配套精细动作标签及个体特异性纹理化虚拟形象,数据量达173,543帧,支持从稀疏关键点到完整表面模型的跨模态研究。其创新性在于将动态3D姿态-形状表征融入动物行为识别任务,为灵长类动物的视觉外观、姿态分析及社交互动研究提供了全新资源。
BigMaQ is the first rhesus macaque action dataset integrating 3D surface modeling, jointly constructed by the University of Tübingen and other institutions. It contains 16-view synchronized videos and skeletal joint rotation data for over 750 interactive scenarios. Collected via 16 calibrated cameras, this dataset is paired with fine-grained action annotations and individual-specific textured virtual avatars, with a total of 173,543 frames, supporting cross-modal research ranging from sparse keypoints to complete surface models. Its innovation lies in integrating dynamic 3D pose-shape representation into animal behavior recognition tasks, providing a brand-new resource for research on visual appearance, pose analysis, and social interaction of primates.
- 1BigMaQ: A Big Macaque Motion and Animation Dataset Bridging Image and 3D Pose Representations蒂宾根大学·Hertie研究所; 马克斯·普朗克智能系统国际研究学院; 鲁汶大学 · 2026年



