Thank you for your contribution. I am trying to reproduce your results but I find the data structure a bit confusing. Your code and the README suggest a data structure that indicates there is only one sequence for training and validation. However, for the KITTI train set, it appears there are 10 different training sequences?
Could you provide some guidance on how to reproduce the results? Should I modify the dataloader for a specific training set, or combine all training sequences into one training set directory (which seems impractical)?
Hi,
Thank you for your contribution. I am trying to reproduce your results but I find the data structure a bit confusing. Your code and the README suggest a data structure that indicates there is only one sequence for training and validation. However, for the KITTI train set, it appears there are 10 different training sequences?
Could you provide some guidance on how to reproduce the results? Should I modify the dataloader for a specific training set, or combine all training sequences into one training set directory (which seems impractical)?