Closed y-kage closed 1 year ago
Q1: Just follow this page Q2: This part is released with different names. Q3: See Q1. We have released the object annotations. Q4: See Q1.
I am able to train now. Sorry for the overlook of that page.
About Q2, could I know which is the file released ? Is it 'config/data/multivideo-mp.yml' ?
Thanks!
Hello @chingswy , thanks for your excellent work!
When I try to follow the tutorial page for the demo "Novel View Synthesis for Human Interaction in Sparse Multi-View Video"
I got this error, It seems that time_offset.json is missing, so I would like to ask what time_offset.json is?
Thank you very much!
You can check the download folder, there must be a file in soccer1_6/output-smpl-3d/time_offset.json
Oh, I see, but the download folder looks outdated, can I have it?
Thank you very much!
I have updated the link.
Thank you again for your help.
I have updated the link.
The link is expired again, can I have it? Thank you very much!
Hello @chingswy , thanks for your wonderful work!
I'm trying to do the demo of 'Novel View Synthesis of Human Interactions From Sparse Multi-view Videos' following the page below using the example data. https://chingswy.github.io/easymocap-public-doc/works/multinb_code.html
However, I get some errors.
When running the motion capture part, I get the the result below, which seems to not working well.
FileNotFoundError: [Errno 2] No such file or directory: 'config/data/mvmp.yml'
[run] none
sh: 1: none: not found
[Skip] Triangulation already done, skipping...
If I skip the motion capture part and do the training, I get the error below. '../data' directory is where the example data is.
The question is similar to #214 ,
Q1. Is it possible to do the demo of Novel View Synthesis of Human Interactions From Sparse Multi-view Videos now?
If so, Q2. Is the 'config/data/mvmp.yml' going to be released? Q3. Is the result of reconstruct the object part correct? Q4. At the training, should '../data/output-smpl-3d/vertices/1/000000.json' be in the data in advance, or is it going to be made in the process of motion capture? Or do I have to do other process before?
Looking forward to your reply! Thanks a lot!