Skip to content

[CVPR'21] [Jittor & Pytorch] DI-Fusion: Online Implicit 3D Reconstruction with Deep Priors

Notifications You must be signed in to change notification settings

huangjh-pub/di-fusion

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 

Repository files navigation

DI-Fusion

This repository contains the implementation of our CVPR 2021 paper: DI-Fusion: Online Implicit 3D Reconstruction with Deep Priors.

Jiahui Huang, Shi-Sheng Huang, Haoxuan Song, Shi-Min Hu

Introduction

DI-Fusion (Deep Implicit Fusion) is a novel online 3D reconstruction system based on RGB-D streams. It simultaneously localizes the camera and builds a local implicit map parametrized by a deep network. Please refer to our technical report and video for more details.

Implementations

We provide two implementations based on Jittor and Pytorch (the latter is also available on the public branch). Please refer to the corresponding folders jittor/ and pytorch/ for specific build and running instructions.

Citation

Please consider citing the following work:

@inproceedings{huang2021difusion,
  title={DI-Fusion: Online Implicit 3D Reconstruction with Deep Priors},
  author={Huang, Jiahui and Huang, Shi-Sheng and Song, Haoxuan and Hu, Shi-Min},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year={2021}
}

About

[CVPR'21] [Jittor & Pytorch] DI-Fusion: Online Implicit 3D Reconstruction with Deep Priors

Topics

Resources

Stars

Watchers

Forks