How close are we to solving vision?

Re:How close are we to solving vision? by mrfusion on Sunday November 13, 2016 @08:51AM (#12943299)

One thing computer vision is missing is making a depth map given a 2d image. You can look at a photograph and describe it as a 3D scene. This will be important for many fields.

[ Parent | on HN ]

Re:How close are we to solving vision? by nicklo on Sunday November 13, 2016 @10:03AM (#12943601)

This problem has already been solved with decent success using deep learning.

See: https://homes.cs.washington.edu/~jxie/pdf/deep3d.pdf

[ Parent | on HN ]

Re:How close are we to solving vision? by mrfusion on Sunday November 13, 2016 @10:29AM (#12943769)

That's a good start. I was thinking you could generate unlimited training data by using a game engine. You'd have the actual 3D model for every single frame.

[ Parent | on HN ]

Re:How close are we to solving vision? by phorese on Sunday November 13, 2016 @12:08PM (#12944299)

Yup, the community is on it!

http://www.cv-foundation.org/openaccess/content_cvpr_2016/ht...

https://link.springer.com/chapter/10.1007/978-3-319-46475-6_...

And there's more every week... Blender, Unity Engine, Unreal Engine, you name it. (Disclaimer: am author on one of these papers)

[ Parent | on HN ]

Re:How close are we to solving vision? by davesque on Sunday November 13, 2016 @12:47PM (#12944532)

I'm aware that certain automotive companies are already doing this.

[ Parent | on HN ]