关于新工作Attention-guided Low-light Image Enhancement的一个问题

[zhangxixi0904]

作者你好！我关注了你光照增强的一个新工作Attention-guided Low-light Image Enhancement，冒昧在这里请教一下。 在合成低光照数据时，你给了两条公式，一条在做伽马变换，一条添加噪声，请问这是两个具有先后顺序的步骤吗？还是完全无关的？因为看到公式中的输入都I_{in}产生了疑惑。

[wxkb]

@zhangxixi0904 您好我最近也在看他新工作的那篇论文，对他噪声图那个计算方式不太了解，方便加个联系方式交流一下吗？2273667502（QQ）

[DineshRajanT]

Hello author! I have followed your new work on illumination enhancement, Attention-guided Low-light Image Enhancement. I take the liberty to ask about it here. When synthesizing low-light data, you gave two formulas, one for gamma transformation and one for adding noise. Are these two sequential steps? Or is it completely unrelated? Because I see the input in the formula I_{in} is confused.

I recently prepared the dataset .....using the formula given in the paper.Yes the steps are sequential.Initially you must apply Gamma and linear transformation followed by adding the noise (poisson in this case) or you also try gaussian ,accordingly.

[DineshRajanT]

Anyone tried to implement the code for "Attention-guided Low-light Image Enhancement"?

[PngYng]

Anyone tried to implement the code for "Attention-guided Low-light Image Enhancement"?

Hi，did you try to implement the code for "Attention-guided Low-light Image Enhancement"? I want to implement too, can you share some experience with me? Thanks

[DineshRajanT]

Anyone tried to implement the code for "Attention-guided Low-light Image Enhancement"?

Hi，did you try to implement the code for "Attention-guided Low-light Image Enhancement"? I want to implement too, can you share some experience with me? Thanks

In the first level I tried its previous version which is MBLLEN. Even though I prepared my own dataset,noise keeps adding at the final output, which is obvious in every encoder-decoder models.Now currently I have the script for synthetic data generation (i.e) paired images.Will try to write the model asap.

[wxkb]

您好，我没有它们的代码，不确定现在作者有没有公布出来，您可以发邮件给作者咨询一下，谢谢！

------------------ 原始邮件 ------------------ 发件人: "Lvfeifan/MBLLEN" <[email protected]>; 发送时间: 2020年10月16日(星期五) 中午11:28 收件人: "Lvfeifan/MBLLEN"<[email protected]>; 抄送: "超人不会飞"<[email protected]>;"Comment"<[email protected]>; 主题: Re: [Lvfeifan/MBLLEN] 关于新工作Attention-guided Low-light Image Enhancement的一个问题 (#4)

Anyone tried to implement the code for "Attention-guided Low-light Image Enhancement"?

Hi，did you try to implement the code for "Attention-guided Low-light Image Enhancement"? I want to implement too, can you share some experience with me? Thanks

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

[Zruto]

Anyone tried to implement the code for "Attention-guided Low-light Image Enhancement"?

Hi，did you try to implement the code for "Attention-guided Low-light Image Enhancement"? I want to implement too, can you share some experience with me? Thanks

In the first level I tried its previous version which is MBLLEN. Even though I prepared my own dataset,noise keeps adding at the final output, which is obvious in every encoder-decoder models.Now currently I have the script for synthetic data generation (i.e) paired images.Will try to write the model asap.

I'm looking forward to your code,thank you

[buptlj]

请问这个新工作的代码有开源的计划吗

[yu-li]

You can find the test code for 'Attention-guided Low-light Image Enhancement' at https://github.com/yu-li/AGLLNet.

[Sanchit231198]

Anyone tried to implement the code for "Attention-guided Low-light Image Enhancement"?

Hi，did you try to implement the code for "Attention-guided Low-light Image Enhancement"? I want to implement too, can you share some experience with me? Thanks

In the first level I tried its previous version which is MBLLEN. Even though I prepared my own dataset,noise keeps adding at the final output, which is obvious in every encoder-decoder models.Now currently I have the script for synthetic data generation (i.e) paired images.Will try to write the model asap.

Hi

Hello author! I have followed your new work on illumination enhancement, Attention-guided Low-light Image Enhancement. I take the liberty to ask about it here. When synthesizing low-light data, you gave two formulas, one for gamma transformation and one for adding noise. Are these two sequential steps? Or is it completely unrelated? Because I see the input in the formula I_{in} is confused.

I recently prepared the dataset .....using the formula given in the paper.Yes the steps are sequential.Initially you must apply Gamma and linear transformation followed by adding the noise (poisson in this case) or you also try gaussian ,accordingly.

Can someone provide some elaboration on how the poission noise is added