search

PeizhuoLi / neural-blend-shapes

An end-to-end library for automatic character rigging, skinning, and blend shapes generation, as well as a visualization tool [SIGGRAPH 2021]
Other
637 stars 92 forks source link

Demonstrate workflow that has all the fingers? #7

Open fire opened 3 years ago

fire commented 3 years ago

Is lacking fingers a limit of the algorithm?

PeizhuoLi commented 3 years ago

Hi, theoretically speaking it is possible to rig fingers though we haven't tried yet. The method itself naturally supports fingers providing a feasible training dataset. But fingers can introduce a lot more burden to the system and it can be computationally expensive to train the system with almost doubled number of joints.

We will release the code for training in the coming week and if you are interested you could try it.

qiji77 commented 3 years ago

你好，我是ucas的研究生，很关注您的工作，想了解下您是否愿意提供训练代码？想尝试一些想法。

PeizhuoLi commented 3 years ago

Hi, the training code has been released. Thanks for your interest!

qiji77 commented 3 years ago

Thank you ！

fire commented 3 years ago

Thank you!

fire commented 3 years ago

I would like to try training all the fingers.

In theory it would be adding the bone effectors to the end of each fingers instead of stopping at the hand.

Question: What are those bone names?

Did I miss anything?

PeizhuoLi commented 3 years ago

Hi, the dataset we use doesn't contain finger's rigging. One possible solution could be switch to SMPL-X model.

TharHtetAungg commented 8 months ago

@fire Did you make it for training finger rig? If so, could you please share about how did you train it?

fire commented 8 months ago

The SMPL-X is licensed for research use only, so I didn't do any work on it.

TharHtetAungg commented 8 months ago

@fire SMPL-X have licensed but it is free to use for their model. Did you test it?