[cs231n_review#Lecture 2-1] Image Classification: A core task in Computer Vision

* The Problem: Semantic Gap

And the computer really is representing the image as this gigantic grid of numbers.

So, the image might be something like 800 by 600 pixels.

And each pixel is represented by three numbers, giving the red, green, and blue values for that pixel.

So, to the computer, this is just a gigantic grid of numbers.

out of this, like, giant array of thousands, or whatever, very many different numbers.

So, we refer to this problem as the semantic gap.

This idea of a cat, or this label of a cat, is a semantic label that we're assigning to this image,

and there's this huge gap between the semantic idea of a cat and these pixel values that the computer is actually seeing.

 

* Challenges: Viewpoint variation

and these pixel values that the computer is actually seeing. And this is a really hard problem because you can change the picture in very small, subtle ways

 

* Challenges: Illumination

* Challenges: Deformation

* Challenges: Occlusion

* Challenges: Background Clutter

* Challenges: Intraclass variation

 

and our algorithms need to be robust to these challenges.