Chongyi Li

Results 63 comments of Chongyi Li

You may change the learning rate to 10-5 or change the value of exposure loss to 0.5 or 0.45. The results are related to the training data. So, if you...

that’s correct. Thanks! 发自我的iPhone > 在 2020年12月25日,下午3:36,Siddharth Shrivastava 写道: > >  > You could replace DCE_net = model.enhance_net_nopool().cuda() in train file by DCE_net = nn.DataParallel(model.enhance_net_nopool()).cuda() > > — >...

请以代码为准,因为论文里面是没有考虑troch.mean的,代码里面是最优的。谢谢! hejing-maker 于2021年1月8日周五 下午2:57写道: > 首先感谢你的开源,我在读代码的时候发现损失函数的权重参数和论文中有些不同,论文中提到:The weights Wcol and WtvA are > set to > 0.5, and 20,但是我看代码中设置为: > Loss_TV = 200*L_TV(A) > > loss_spa = torch.mean(L_spa(enhanced_image, img_lowlight)) > >...

Yes, it is. We tested the runtime of Zero-DCE. Please see Table 3 of our paper. xuhui1994 于2020年12月1日周二 下午1:59写道: > I wanna to know can this code real-time to product...

Too busy to find the source testing code. I will check it out soon. 发自我的iPhone > 在 2020年12月25日,下午3:51,Siddharth Shrivastava 写道: > >  > Any updates on this? Would be...

Hi, There is no natural difference. You can just use ^2, but you have to change the weight of color loss in the training phase. BEST, Bo Sun 于2020年7月25日周六 下午10:09写道:...

Hi, you can find the PI metric here: https://github.com/chaoma99/sr-metric. kingloser 于2020年7月9日周四 下午7:00写道: > Sorry to bother you, I'd like to ask you the User study (US)↑/ Perceptual > index (PI)...

可以使用 Matlab 里面的 heatmap 命令,或者Python matplotlib 命令。 航海的W ***@***.***> 于2021年11月23日周二 下午4:30写道: > 作者您好!请问下论文内 Figure3用于可视化的 b、c、d热力图代码可以分享一下吗? > > — > You are receiving this because you are subscribed to this thread....

Its transmission map generated by GDCP, not depth map. It is almost impossible to obtain the depth of an underwater image. I am not sure if your GDCP code is...

Please check my Notice. Thanks.