tryink
tryink
> > > Thanks for your interest, we use the StyleGAN inversion checkpoint (https://drive.google.com/file/d/1bMTNWkh5LArlaWSc_wa8VKyq2V42T2z0/view). Note that you need to align the faces before inversion. > > > > > >...
> > > > Hello, I am new to this area, can you give me some advice on how to use FSLSD code for preprocessing and training Hi, I just...
> Hello, I meet the same question. Have you solved it? Which checkpoint do you use in pSp? What code in pSp do you take, `encoder.code` or `encoder.code + pSp.avg_latent`...
Hi, I have confirmed the preprocessing steps and tried again, but just got the same meaningless results. And I’m still looking for where I did wrong.
感谢回复！这样的话也就是说代码环境对cosface的效果影响很大？以及insightface-repo的人脸识别模型是arcface，和文中id_embedding用到的是同一个，可以用相同的模型作为评价指标吗？ ------------------ 原始邮件 ------------------ 发件人: "neuralchen/SimSwap" ***@***.***>; 发送时间: 2022年7月29日(星期五) 中午11:01 ***@***.***>; ***@***.******@***.***>; 主题: Re: [neuralchen/SimSwap] id (Issue #305) 哥们, 我的建议是用insightface。 @neuralchen 论文里放的cosface github链接里面的cosface版本用到的python是2点几的版本... — Reply to this email directly, view it on GitHub, or...
您好， 好的，感谢您的回复。 看来环境对模型的影响还是挺大的，请问您有尝试过用torch1.几复现cosface吗？如果有的话，复现效果如何呢？ 我也不一定非要完全复现论文，更多的是想得到合理的指标测试模型，Insightface用起来确实要方便一些，但我疑惑的点在于，在您看来，如果人脸合成过程用到了arcface提取身份特征，测试时再继续用arcface计算id_retrieval，是否会导致结果偏差呢？另外，或者如果考虑到训练集和测试集不相关（来自两个数据集），这种偏差还存在吗？我实在是想不明白，期待您的回复，非常感谢！ 祝好！ ------------------ 原始邮件 ------------------ 发件人: "Marco ***@***.***>; 发送时间: 2022年7月31日(星期天) 凌晨0:57 收件人: ***@***.***>; 抄送: ***@***.***>; ***@***.***>; 主题: Re: [neuralchen/SimSwap] id (Issue #305) 并不是这么说, 只是我个人嫌重新搭环境麻烦而已XD 也许我的情况和你的不太一样, 如果是要完全复现论文的话, 那还是用cosface吧 — Reply...
> > > 当然，芽，今晚或早上将发布。你放弃什么？你有什么问题？ > > > > > > Bro, did you post your update? I'm excited to see it > > Posting today!! Once I get home :)...
> Your understanding is correct. Note that the number of "samples" here is comparable to other DDPM on the same dataset. Maybe, the number of epochs is not interpreted the...
> @lmm077 Hello, thank you for your attention to DreamVideo. Ideally, we plan to open source the code in about one to two months. We need to go through the...
> Sorry, but this work was done during my internship at Tencent AI Lab. Since I have left the company, I can no longer access the data files stored on...