Open karthik101200 opened 2 months ago
Hi, it is definitely possible to use other dataset. I think you need to write your own data loader to load the training images, poses and initial point cloud. You can check here for reference https://github.com/autonomousvision/gaussian-opacity-fields/blob/main/scene/dataset_readers.py#L226-L260
Hi thank you for the work. I want to ask if I can bypass COLMAP poses and instead use localization data that I already have from ROS or any other method. I understand this work may not be related or considered. But are there suggestions and markers to code where I can change it