search

syguan96 / DynaBOA

[T-PAMI 2022] Out-of-Domain Human Mesh Reconstruction via Dynamic Bilevel Online Adaptation
225 stars 19 forks source link

Could you show the final tree of human3.6m dataset according to your processing code? #28

Open Mirandl opened 2 years ago

Mirandl commented 2 years ago

Hi, your work's so marvellous and I'm trying to realise your code. cause I have this processed dataset early before and I wonder if my data structure is totally the same as yours. So It will help me very very a lot if you can share the structure of your processed human3.6m dataset! Thank you and waiting for your reply

syguan96 commented 2 years ago

Hi @Mirandl , thanks! You can refer to this script to see how I process Human3.6M.

Processed file looks like: image

Mirandl commented 2 years ago

Hi, thank you very much for your sharing! I notice that there's some difference in the way the dataset is handled between your BOA and DynaBOA. I mean that I notice the BOA's more abundant. Is it matter which way I choose to handle it?

syguan96 commented 2 years ago

It doesn't matter. Human 3.6M is used to train the source model. During training, you need a bbox to crop input image, and SMPL annotation and 3D/2D keypoints as supervisions.

Mirandl commented 2 years ago

okay I got it! Thank you so much~