How to run it on kitti dataset

[Mr-ind1fferent]

Thanks for the amazing work!

I want to know how to run mvsgaussian on autonomous driving dataset such as kitti

[TQTQliu]

Thanks!
I'm not familiar with kitt dataset, can you describe how the dataset is organized?

[Mr-ind1fferent]

This is the structure of kitti data set 2011_09_26_drive_0001, including the data collected by four cameras and the corresponding time, lidar point cloud, internal and external parameters of the cameras, etc. How can I convert it into the input format required by mvsgaussian for training?

[Mr-ind1fferent]

This is the structure of kitti data set 2011_09_26_drive_0001, including the data collected by four cameras and the corresponding time, lidar point cloud, internal and external parameters of the cameras, etc. How can I convert it into the input format required by mvsgaussian for training?

This is the data I have processed.

[TQTQliu]

Since the kitti dataset provides the image and camera parameters, one direct solution is to reconstruct sparse model from known camera poses to obtain colmap-format data, and then run mvsgaussian.

[Mr-ind1fferent]

Is it not suitable for multi camera as input