It finds small visual clues first and then combines them to identify the object.
Example
You show a CNN 10,000 pictures of cats and dogs.
After training, when you show it a new picture, it might say:
"I see pointy ears, whiskers, and cat-like eyes. 98% chance this is a cat."
Convolutional Neutral Network.
CNN is a type of AI specially designed to recognize patterns in images by scanning small parts of an image and gradually building an understanding of what's there. This makes CNNs very effective for image recognition, facial recognition, medical.
This makes CNNs very effective for image recognition, facial recognition, medical scans, self-driving cars, and similar tasks.
Simple analogy
Think of a jigsaw puzzle:
First you find corner pieces.
Then edges.
Then larger sections.
Finally you see the whole picture.
@gemskrl @Phatymarh44 Some people like to cause trouble by insulting other people's religion. Must you speak if you don't like her opinion. Stop insulting our prophet
@tebogosmomo@markoscar_@aime_mwizaa That is why we avoid such women. When a woman is financially stable she looks out for her self. But a financially stable man looks out for his Family. You see the difference