Thanks for the great work! I'm testing the model on several datasets. It performs well on GSO data but when I tested it on my customized dataset somehow I got very blurry results like this:
I guess it might have something to do with the camera parameters. I can get better results when scaling down camera positions, so I'm wondering is there a optimized setting for camera parameters, such as cam distance, near_far range, etc.?
Thanks for the great work! I'm testing the model on several datasets. It performs well on GSO data but when I tested it on my customized dataset somehow I got very blurry results like this:
I guess it might have something to do with the camera parameters. I can get better results when scaling down camera positions, so I'm wondering is there a optimized setting for camera parameters, such as cam distance, near_far range, etc.?