Hacker News new | ask | show | jobs
by eps 239 days ago
Having the skeleton be auto-deduced (through the ML or otherwise) would've been a very natural fit here as a feature. Knowing that it's basically possible to do, selecting the skeleton type and then adjusting it manually, joint by joint, seems ... archaic?
1 comments

Have you done 3D modelling/rigging/animating yourself before?
Why the condescending tone, OP has a point. Techniques like RigAnything[0] exist - whether this is within the scope of this project is another topic, but this seems like a reasonable request to me (especially considering how many x_2_y-named projects are ML related)

[0] https://www.liuisabella.com/RigAnything/

This kind of work is way more difficult and involved than it appears to be. Sure, you can auto-rig stuff, but it won't be perfect and will require tweaking. Rigging alone is also very far from a complete result