diviswen/Cycle4Completion

How to train on custom complex data?

Opened this issue · 0 comments

Hi has anyone been able to train it on their own custom dataset? If so, which categories and how well did it perform? Which part of the code needs to be modified? I want to train it to do completion of dense pointclouds of individual trees. Do you think it will work well and how many trees will I need? Can I train it from scratch or will I need to use transfer learning?