Hello,
I'm interested in running your code (just for the inference) on my custom dataset .
However, the custom dataset you describe seems to be constructed with 3D meshes, while my dataset consists only of RGB frames(large-scale scene, like street view).
I have the camera pose for each frame in the form of a 4x4 matrix. You've provided very clear instructions, but I'm not sure how to proceed with testing in my case. I'm unsure how to handle the 'settings.json' and 'occupied_pose.pt' files with my rgb frames and camera poses.
I would be grateful if you could guide me on how to test with my type of dataset.
Thank you for your assistance.
Hello,
I'm interested in running your code (just for the inference) on my custom dataset .
However, the custom dataset you describe seems to be constructed with 3D meshes, while my dataset consists only of RGB frames(large-scale scene, like street view).
I have the camera pose for each frame in the form of a 4x4 matrix. You've provided very clear instructions, but I'm not sure how to proceed with testing in my case. I'm unsure how to handle the 'settings.json' and 'occupied_pose.pt' files with my rgb frames and camera poses.
I would be grateful if you could guide me on how to test with my type of dataset.
Thank you for your assistance.