How is machine learning used here? (the Professor mentioned that it was). I've never thought of solving inverse kinematics problems as ML, unless this was some approximate solution estimate thingy...
P.S. You can also tell this is old by the browser GUI lol
geos98
@rsha256, my understanding of the paper is the following. The paper is not solving IK using ML, but rather using ML to determine a better solution or rather a solution that fits a certain "style" (remember that an IK could have multiple solutions).
The paper is from 2004, it is very old. It was so old that when they say ML, they are not referring to a neural net based ML.
CardiacMangoes
Do the inverse kinematics for a whole humanoid figure contain a physics-based bias? I imagine you could estimate the center of mass for a figure to help learn the position of the figure as done here.
How is machine learning used here? (the Professor mentioned that it was). I've never thought of solving inverse kinematics problems as ML, unless this was some approximate solution estimate thingy...
P.S. You can also tell this is old by the browser GUI lol
@rsha256, my understanding of the paper is the following. The paper is not solving IK using ML, but rather using ML to determine a better solution or rather a solution that fits a certain "style" (remember that an IK could have multiple solutions).
The paper is from 2004, it is very old. It was so old that when they say ML, they are not referring to a neural net based ML.
Do the inverse kinematics for a whole humanoid figure contain a physics-based bias? I imagine you could estimate the center of mass for a figure to help learn the position of the figure as done here.