-
Notifications
You must be signed in to change notification settings - Fork 119
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
How is the effect of training? #17
Comments
@zhijian-liu @YuliangXiu @Vera-He @songzhaozhe |
I get a bad result on scannet about 84.7 use your code strictly, but your paper's report is 86. |
@meteorshowers 你的结果怎么样?不是这种数字,是预测的可视化效果怎么样?我觉得这数字上的差异毕竟不是太大可以接受的。毕竟我们只是局部最优解,他可能是训练了很多次找到了更好的局部最优解。你也可以多训练几次,也许会有提升。 |
可以@下作者 |
@meteorshowers 训练的时候只需要像素的坐标点就可以了吗,不需要像素值吗?我看到代码里好像没有用到像素值 |
@Laviyy 点云的就是点的坐标位置。你理解的像素值是什么?他其实是提取每个点和周围点的特征,让其具有局部特征,以弥补pointnet的不足。 |
Is there code implemented by Pytorch? |
Could you provide a trained model?
The text was updated successfully, but these errors were encountered: