serdarch / SERNet-Former

[CVPR 2024 Workshops] SERNet-Former: Semantic Segmentation by Efficient Residual Network with Attention-Boosting Gates and Attention-Fusion Networks
Other
39 stars 3 forks source link

code release #3

Closed shaoyandea closed 3 months ago

shaoyandea commented 3 months ago

your work is great! can you release code , i am very interesting in it

serdarch commented 3 months ago

Many thanks for your comments and interests. I have just committed some updates to the codes and added new examplar codes. Please find them in the repo.

Please also ensure that the original model is being developed in MATLAB. Thus, the code is fundamentally different from PyTorch. So, this repository is only being developed for additional language support and as a backup repository to the network with many basic and tutorial-based codes to emulate the development progress of the network.

I hope this answers your inquiry. Best

serdarch commented 3 months ago

This repository is only being developed for additional language support and as a backup repository to the network with many basic and tutorial-based codes to emulate the development progress of the network.

I hope this answers your inquiry. Best

shaoyandea commented 3 months ago

thank you for your reply,  I am doing some research about semantic segmentation and I believe this code and paper will give me great help in research, hoping for  complete code release .

I'm sorry if I caused you any trouble.

Best 

------------------ 原始邮件 ------------------ 发件人: "serdarch/SERNet-Former" @.>; 发送时间: 2024年6月28日(星期五) 凌晨1:17 @.>; @.**@.>; 主题: Re: [serdarch/SERNet-Former] code release (Issue #3)

This repository is only being developed for additional language support and as a backup repository to the network with many basic and tutorial-based codes to emulate the development progress of the network.

I hope this answers your inquiry. Best

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>