Further to this:
jalopnik.com:Apple's Secretive Car Program Might Have A Way For Autonomous Cars To Spot 3D Objects
Inside the dotted orange line is what they call Voxelnet. The leftward magenta boxes use (I suspect) something like OpenVDB, that I've
spoken of before. Creating a sparse, shallow inverted B-tree allows the extraction of the point-wise features. It essentially
is the sparse 4D tensor field they talk about.
Quote:
|
Originally Posted by jamesqf
The problem with that approach is that if you make your computing units into neuromorphic "brainz", there's no reason to think that they won't acquire all the flaws* of biological brains in the process.
|
Exponential autism, curse or menace?.
https://slashdot.org/