Lecture 27: Inverse Graphics/Research In Angjoo's Group (8)
yang991178
The reduction of 3D objects into 2D reminds of the similar reduction into visible color space by human eyes. However, it is obviously impossible to trick brains in a way similar to color cognition by restrict human perception to 2D. :(
sethzhao506
It's easy to see that when we observe the 2D mapping from 3D objects it's hard to know what would the object exactly be like. Reconstructing 3D object then requires other aspects of data.
ZizhengTai
I learned in CS 294-137 last semester that humans deal with this with a lot of different mechanisms. For example, our two eyes capture different images, and the brain can use the differences between them to infer what the 3D object looks like. The brain also uses the changes in captured images when we look at an object from slightly different angles to infer its 3D properties.
The reduction of 3D objects into 2D reminds of the similar reduction into visible color space by human eyes. However, it is obviously impossible to trick brains in a way similar to color cognition by restrict human perception to 2D. :(
It's easy to see that when we observe the 2D mapping from 3D objects it's hard to know what would the object exactly be like. Reconstructing 3D object then requires other aspects of data.
I learned in CS 294-137 last semester that humans deal with this with a lot of different mechanisms. For example, our two eyes capture different images, and the brain can use the differences between them to infer what the 3D object looks like. The brain also uses the changes in captured images when we look at an object from slightly different angles to infer its 3D properties.