Training+finetuning configurations
gallif opened this issue · comments
gallif commented
Hi, first of all great work and fantastic code!
I'm trying to recreate your reported results on the Sintel Training:
- Did you train and evaluate using the complete Sintel train dataset?
- I noticed that occlusion transform (run_ot) is set to false in the ar config file. Was it used during finetuning?
- Are there other parameters I should consider?
Thanks!
Liang Liu commented
- Yes, I did. The results in Table 1 were trained on the complete Sintel train dataset.
- Since OT significantly increases training time, I did not use it in most of ablations, but it can still bring performance gains(refer to Table 4). The results in Table 1 used all kinds of transformations.
- You can reproduce the results just use the default parameters.
gallif commented
Thank you for the reply.
I will do as you suggested.