aim-uofa/frozenrecon

[ICCV2023] 🧊FrozenRecon: Pose-free 3D Scene Reconstruction with Frozen Depth Models

frozenrecon's Issues

How to generate red triangulations in reconstructed scene like your homepage?

Hello,i am very grateful for your work. I want to know how to generate red triangulations which represent pose in reconstructed scene in your github code. I want to visualize pose. I am grateful for your soon response.

在移动范围小的场景下，重建结果会成为一个平面

感谢您非常好的工作，当在移动范围小的场景下，重建结果会成为一个平面。请问您知道这是什么原因导致的吗？

When the code will be public?

Thanks for your great jobs!
I would like to use this project in my paper, and you said the code will be public in the ICCV paper. Can you please kindly let us know when the code will be public?

Can I output the pose with an absolute scale?

Dear author，
Thanks for your great work, but I have a problem with this, I want to predict the pose from the video. Can I output the pose with an absolute scale?

Question about world coordinate systems

Hi! Thanks for the great work!

I'm trying to test on my own data and input the gt camera intrinsic and extrinsic, but I don't see the transformation in the provided dataloader, so how do I define the world coordinate systems? (I'm trying with w2c (world to camera) in OpenCV/COLMAP system, but seems that it cannot work well.)