Skip to content

Building Change Detection using Cross-temporal Feature Interaction Network (ICASSP 2023)

Notifications You must be signed in to change notification settings

ZhengJianwei2/CTFINet

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

30 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CTFINet

This is the code for "Building Change Detection using Cross-temporal Feature Interaction Network", which was accepted by ICASSP 2023. Based on a naive backbone (ResNet18) without sophisticated structures, our model outperforms other state-of-the-art methods on three datasets in terms of both efficiency and effectiveness. If you you encounter any doubts, please fell free to concact us: fyc@zjut.edu.cn or zjw@zjut.edu.cn.

Visual results

Visualization of feature maps taking bi-temporal RS images from LEVIR-CD as example.

Architecture of the proposed CTFINet

Quantitative results

Comparison results on detection of building changes.

About

Building Change Detection using Cross-temporal Feature Interaction Network (ICASSP 2023)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published