Re:Nvidia on new self-driving system: “basically 5 years ahe...
by pakl on Sunday November 13, 2016 @12:25AM
(#12942002)
> There are too many hard cases for vision.
Even worse, most machine learning vision approaches make the vision problem much harder on themselves. They do not treat the visual world as the dynamic physical interacting processes that give rise to it.
A disadvantage treating of vision as static frames of pixels is that deep feedforward networks have to "memorize" all the physical dynamical effects of e.g., shadows on textures.
Such systems cannot generalize well. A more promising approach is hierarchical systems with ubiquitous recurrent connectivity.
Even worse, most machine learning vision approaches make the vision problem much harder on themselves. They do not treat the visual world as the dynamic physical interacting processes that give rise to it.
A disadvantage treating of vision as static frames of pixels is that deep feedforward networks have to "memorize" all the physical dynamical effects of e.g., shadows on textures.
Such systems cannot generalize well. A more promising approach is hierarchical systems with ubiquitous recurrent connectivity.