Closed 16619884724 closed 4 years ago
yifanfeng97
commented
4 years ago Thank you for your attention. For the used feature: we use the feature map that just before the first fully connected layer. For source code: Unfortunately, I'm so busy recently (especially the two weeks). Fortunately, I plan to release the verified GVCNN source code soon.
16619884724
commented
4 years ago OK,It is so kind of you.Looking forward to your source code and next work.Thanks again for your reply
------------------ 原始邮件 ------------------ 发件人: "Yifan Feng(丰一帆)"<notifications@github.com>; 发送时间: 2019年12月10日(星期二) 下午2:34 收件人: "iMoonLab/HGNN"<HGNN@noreply.github.com>; 抄送: "Confession"<1773507074@qq.com>;"Author"<author@noreply.github.com>; 主题: Re: [iMoonLab/HGNN] features (#12)
Thank you for your attention. For the used feature: we use the feature map that just before the first fully connected layer. For source code: Unfortunately, I'm so busy recently (especially the two weeks). Fortunately, I plan to release the verified GVCNN source code soon.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.
yifanfeng97
commented
4 years ago Done! I'll close it.
Hello, your work is excellent, and I am very interested in it. Do the extracted GVCNN features you used refer to the final shape descriptor that has performed grouping and inter-group fusion? I know that GVCNN is also your work, but there is no relevant open source code. Would you please provide some code? Thank you