D-IF: Uncertainty-aware Human Digitization via Implicit Distribution Field
Xueting YangYihao LuoYuliang XiuWei WangHao XuZhaoxin Fan
Xueting YangYihao LuoYuliang XiuWei WangHao XuZhaoxin Fan
0被引用
2笔记
开学季活动火爆进行中，iPad、蓝牙耳机、拍立得、键盘鼠标套装等你来拿
摘要原文
Clothed human reconstruction is an important but challenging task for many applications, including augmented reality, virtual reality and metaverse. The use of deep implicit function sparks a new era of image-based 3D clothed human reconstruction. The vast majority of works locate the implicit surface by regressing the deterministic perpoint implicit value. However, should all the points, like near-surface and floating points, be treated equally? In this paper, we replace the implicit value with adaptive uncertainty distribution, to differentiate the points located at different distance fields. Such a simple "value ⇒ distribution" transition, finally leads to a significant improvement on almost all the baseline approaches. Qualitative results show that the models, which are trained with our uncertainty distribution loss, could recover more detailed wrinkles, and human-like limbs.
