This website requires JavaScript.

Hi-LASSIE: High-Fidelity Articulated Shape and Skeleton Discovery from Sparse Image Ensemble

Chun-Han YaoWei-Chih HungYuanzhen LiMichael RubinsteinMing-Hsuan YangVarun Jampani
Dec 2022
摘要
Automatically estimating 3D skeleton, shape, camera viewpoints, and partarticulation from sparse in-the-wild image ensembles is a severelyunder-constrained and challenging problem. Most prior methods rely onlarge-scale image datasets, dense temporal correspondence, or human annotationslike camera pose, 2D keypoints, and shape templates. We propose Hi-LASSIE,which performs 3D articulated reconstruction from only 20-30 online images inthe wild without any user-defined shape or skeleton templates. We follow therecent work of LASSIE that tackles a similar problem setting and make twosignificant advances. First, instead of relying on a manually annotated 3Dskeleton, we automatically estimate a class-specific skeleton from the selectedreference image. Second, we improve the shape reconstructions with novelinstance-specific optimization strategies that allow reconstructions tofaithful fit on each instance while preserving the class-specific priorslearned across all images. Experiments on in-the-wild image ensembles show thatHi-LASSIE obtains higher quality state-of-the-art 3D reconstructions despiterequiring minimum user input.
展开全部
图表提取

暂无人提供速读十问回答

论文十问由沈向洋博士提出,鼓励大家带着这十个问题去阅读论文,用有用的信息构建认知模型。写出自己的十问回答,还有机会在当前页面展示哦。

Q1论文试图解决什么问题?
Q2这是否是一个新的问题?
Q3这篇文章要验证一个什么科学假设?
0
被引用
笔记
问答