BriefGPT.xyz
Nov, 2020
FlowStep3D: 模型展开的自监督场景流估计
FlowStep3D: Model Unrolling for Self-Supervised Scene Flow Estimation
HTML
PDF
Yair Kittenplon, Yonina C. Eldar, Dan Raviv
TL;DR
该论文提出了一种基于循环神经网络的场景流场估计方法，并通过迭代逐步优化的方式提高了其预测精度，在FlyingThings3D数据集上训练后成功地将其推广到实际应用中，并在KITTI基准测试中大幅优于现有方法。
Abstract
Estimating the
3d motion
of points in a scene, known as
scene flow
, is a core problem in computer vision. Traditional learning-based methods designed to learn end-to-end 3D flow often suffer from poor generalizat
→