Skip to content
/ crackmer Public

Complementary networks based on CNN and Transformer

Notifications You must be signed in to change notification settings

zZhiG/crackmer

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

43 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Dual-path network combining CNN and transformer for pavement crack segmentation

Main framework of the proposed network is as follows:

We have uploaded a model code implemented using PyTorch, which is simple and brief. We also uploaded the example weights trained on the DeepCrack dataset, which can help us better reproduce the model's performance.

Here are two visual examples:

(1) Obtain binary segmentation mask:

(2) Obtain the segmentation mask based on the original image:

More details will be described in our paper. If this work is helpful to you, or if you need to use our network in your work, please cite us:

@article{WANG2024105217,
title = {{Dual-path network combining CNN and transformer for pavement crack segmentation}},
journal = {Automation in Construction},
volume = {158},
pages = {105217},
year = {2024},
issn = {0926-5805},
doi = {10.1016/j.autcon.2023.105217},
author = {Jin Wang and Zhigao Zeng and Pradip Kumar Sharma and Osama Alfarraj and Amr Tolba and Jianming Zhang and Lei Wang}
}

About

Complementary networks based on CNN and Transformer

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages