xuchen-ethz / continuous_view_synthesis Goto Github PK

hi,
thank your great work, can you upload the evaluation code? thanks.

Hi,could you explain why the relative pose is calculated as RA.T@RB and T = RA.T.dot(TB-TA)/50 when processing the kitti dataset?
i got the wrong result with other calculation formulas.
thank you!

Thanks to the open source code, I have some questions about the demo test. When I use demo_kitti.py , the value of event.xdata will show None while the mouse is moving. Can you tell me why that is.

Thank you for your open source code. The following error occurs when I run demo_car.py，

AttributeError: 'scipy.spatial.transform.rotation.Rotation' object has no attribute 'as_dcm'

Can you tell me the correct version of scipy

Hi, sir:
What is the unit of geometric transformation using in view point change?
Thanks.

Question - What actually are the first three values in each entry in poses.txt? According to the official repo (https://github.com/bostondiditeam/kitti/blob/master/Papers_Summary/Geiger2013IJRR/readme.md), KITTI odometry has 7 values (the first three being box size along respective dimensions), but in poses.txt there're just six... And why are they involved in the camera rotation matrix?

I would really appreciate it If you could explain what is going on here:

Thank you

there is a request about the dataset of car, would you like to tell me the number of epoch about car dataset? thank you.

Hi,

I see that you are dividing the translation vector by 50( see link below). Do I have to modify my translation vectors? Could you please explain the way to represent the camera poses for street scene datasets?

I have studied your article named recently. Could you please provide TXT files of training data and test data.Thanks for your reply.

您好，非常感谢您的杰出贡献！我想请教一下，我看论文中写道是：通过透视投影从解码器所预测的深度图Dt来得到 dense backward flow map Ct→s，这个过程好像需要camera intrinsic matrix K，请问这个K是怎么得到以及确定的呢？如果换一个数据集，这个K不需要作出任何改变吗？如何根据数据集来进行改变呢？（我看代码中的kitti数据集与car数据集的intrinsics就是不一样的矩阵）

真心期待得到您的解答！