Skip to content

Latest commit

 

History

History
30 lines (22 loc) · 1.38 KB

README.md

File metadata and controls

30 lines (22 loc) · 1.38 KB

This is the code repository for "Hierarchy Denoising Recursive Autoencoders for 3D Scene Layout Prediction" .

IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2019

Created by Yifei Shi, Angel Xuan Chang, Zhelun Wu, Manolis Savva and Kai Xu

teaser

Usage - VDRAE

Dependancy

The code includes two components: data preprocessing and VDRAE.

Data preparation

You need to process the raw data of S3DIS or Matterport3D before feed it to VDRAE, please see preprocess for details.

VDRAE

The VDRAE is the main network of this project. After data preparation, you can train and test the scene by looking at vdrae.

Citation

If you find the code is useful, please cite:

@inproceedings{shi2019hierarchy, 
author = {Yifei Shi and Angel Xuan Chang and and Zhelun Wu and Manolis Savva and Kai Xu}, 
booktitle = {Proc. Computer Vision and Pattern Recognition (CVPR), IEEE}, 
title = {Hierarchy Denoising Recursive Autoencoders for 3D Scene Layout Prediction}, 
year = {2019}
}