BriefGPT.xyz
May, 2019
从单张图像学习重构3D曼哈顿线框
Learning to Reconstruct 3D Manhattan Wireframes from a Single Image
HTML
PDF
Yichao Zhou, Haozhi Qi, Yuexiang Zhai, Qi Sun, Zhili Chen...
TL;DR
该论文提出了一种从单个图像中获取紧凑且精确的3D线框表示的方法，通过有效地利用全局结构规律，该方法训练卷积神经网络以同时检测显著的交叉点和直线，并预测它们的3D深度和消失点, 并通过平行结构先验进一步重构完整的3D线框模型。
Abstract
In this paper, we propose a method to obtain a compact and accurate
3d wireframe representation
from a single image by effectively exploiting global structural regularities. Our method trains a
convolutional neural netw
→