News for hackers. Stuff that matters.
Re:Where we see shapes, AI sees textures by personjerry on Monday July 08, 2019 @12:58PM (#20384375)
This shouldn't be news. Most computer vision NNs are CNNs where the convolutions are basically translating the image into small "textures". "AI sees textures" yeah because that's how we present data to it!
Re:Where we see shapes, AI sees textures by hammock on Monday July 08, 2019 @02:43PM (#20385509)
I wonder what would happen if we fed it SVGs instead of JPGs. (Shape files)
Re:Where we see shapes, AI sees textures by tastroder on Monday July 08, 2019 @02:49PM (#20385577)
At that point it's less a task in the context of image processing and more one of graph learning / language processing, depending on how you want to formulate it.
Re:Where we see shapes, AI sees textures by wongarsu on Monday July 08, 2019 @03:19PM (#20385827)
Human vision certainly contains a component where we do 3d reconstruction and adjust our perception based on that. Many optical illusions depend on this step.

If we want more human like machine vision then having passes in image processing that deal with more abstract data sounds like a great idea

Re:Where we see shapes, AI sees textures by mirekrusin on Monday July 08, 2019 @02:54PM (#20385622)
I don't think those networks can work with variable inputs, they use fixed size images, svg describes paths that highly vary.
Re:Where we see shapes, AI sees textures by abrichr on Monday July 08, 2019 @03:05PM (#20385732)
That's true for convolutional networks. However, recurrent architectures aren't similarly limited, e.g SketchRNN:

https://github.com/tensorflow/magenta/blob/master/magenta/mo...