Relative Pose Estimation for RGB-D Human Input Scans via Human Completion
文献类型:会议论文
作者 | Pengpeng Liu1,2; Guixuan Zhang1; Hu Guan1; Jie Liu1; Shuwu Zhang1; Zhi Zeng1 |
出版日期 | 2021-11 |
会议日期 | November 18-21, 2021 |
会议地点 | Beijing, China |
国家 | 中国 |
英文摘要 | Relative pose estimation for human scans enjoys a promising prospect. However, most existing methods mainly focus on indoor or outdoor scenes, requiring considerable overlap between the inputs. We present a technique for estimating the relative pose whatever the overlap between the human RGBD input scans is. For non-overlapping scans, the insight is to take advantage of the underlying human geometry prior as much as possible. We utilize the implicit function model for human reconstruction, enriching abundant hidden cues for unseen regions, then we use the completed human geometry to get a stable pose estimation. Our evaluation shows that our approach outperforms considerably than standard pipelines in non-overlapping setting, without compromising performance over overlapping input scans. |
源文献作者 | 中国科学院自动化研究所,中国传媒大学 |
产权排序 | 1 |
源URL | [http://ir.ia.ac.cn/handle/173211/47517] |
专题 | 数字内容技术与服务研究中心_新媒体服务与管理技术 |
通讯作者 | Zhi Zeng |
作者单位 | 1.Institute of Automation, Chinese Academy of Sciences 2.Univeisity of Chinese Academy of Sciences |
推荐引用方式 GB/T 7714 | Pengpeng Liu,Guixuan Zhang,Hu Guan,et al. Relative Pose Estimation for RGB-D Human Input Scans via Human Completion[C]. 见:. Beijing, China. November 18-21, 2021. |
入库方式: OAI收割
来源:自动化研究所
其他版本
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。