返回榜单

FORTH-ModelBasedTracker / MocapNET

一种实时方法,能够直接在流行的Bio Vision Hierarchy(BVH)格式中估算3D人体姿势,该方法基于来自单目彩色图像的2D身体关节信息。我们的贡献包括:(a) 一种新颖且简洁的2D姿势NSRM表示方式;(b) 一个人体姿态分类器以及一组经过姿态调整的模型

仓库 Star
956总榜 #41,449
Fork
144
最近推送
2 个月前更新
仓库创建
2019 年创建
应用与完整系统通用C++许可证未声明
查看源码官网

用途与标签

项目原文 · GitHub
A real-time method that estimates the 3D human pose directly in the popular Bio Vision Hierarchy (BVH) format, given estimations of the 2D body joints originating from monocular color images. Our contributions include: (a) A novel and compact 2D pose NSRM representation. (b) A human body orientation classifier and an ensemble of orientation-tuned

用之前确认

  • 没有声明许可证:商用和分发条件无法确认,使用前请联系作者。

Star 走势

95695795711-06 · 95610-0810-1310-1810-2310-2811-0211-06

首次收录 2026-11-06采集时间 2026-11-06 08:36 UTC数据来源许可证原文

介绍根据项目公开资料整理,未经实际测试。 · GitHub 仓库 Star,不是独立使用人数 · 商用三档依据许可证元数据自动分类;具体义务需结合实际使用方式核对。未知不等于禁止商用。