paint-brush
How Do Artificial Neural Network Recognize Imagesby@priyanshujain
1,917 reads
1,917 reads

How Do Artificial Neural Network Recognize Images

by PRIYANSHU JAINMarch 10th, 2017
Read on Terminal Reader
Read this story w/o Javascript
tldt arrow

Too Long; Didn't Read

ANN is something rough mathematical cartoon of how a <a href="https://hackernoon.com/tagged/biological" target="_blank">biological</a> neural <a href="https://hackernoon.com/tagged/network" target="_blank">network</a> works. In biological brain we have individual cells called neurons, each neuron looks what other neuron looks at to what it’s neighbor has to say then it decides what it wants to say. In artificial neural network we have little mathematical functions we put them in some organised structure and then we say okay! you guys are all together to learn to do this task.

Company Mentioned

Mention Thumbnail
featured image - How Do Artificial Neural Network Recognize Images
PRIYANSHU JAIN HackerNoon profile picture

ANN is something rough mathematical cartoon of how a biological neural network works. In biological brain we have individual cells called neurons, each neuron looks what other neuron looks at to what it’s neighbor has to say then it decides what it wants to say. In artificial neural network we have little mathematical functions we put them in some organised structure and then we say okay! you guys are all together to learn to do this task.

There are lots of neural nets that are really great at going and recognizing, you know this is a cat, this is a dog, this is a frog, this is a mouse and things like that. These neural nets works in multiple layers so this kind of machine learning is also called deep learning. Neuron at bottom layer do tiny little work like looking at a piece of a picture and making some computations about it, it does not understand much specifically but what does neuron does understand is that I am giving a signal that is useful to somebody who is giving a signal , who is giving a signal , who is giving a signal , who is giving a signal and so on. They are kind of able to unfurl this really high dimensional knot and pull it apart and make it easier to go in separate different things that are close together on the surface from things that are were tangled all together earlier. But then a top we put two neurons and these neurons look at the whole picture so far. They are basically experts at making the final call figuring out all the layers below me said these things so I know that this is a cat or at least I am 92% sure that this is a cat so it’s basically a cat.

As in the photo below we can say that right image in right is 92% likely to be of a man.

But it takes a long time to learn you show it a picture of a car when it’s early in learning, the very next time you show it a picture of a car it’s only a little bit more likely to say it’s a car. It does not get in even though that was the very last thing it saw. Where as you say to a kid, that’s an filing cabinet and then a second later you say what’s that? He’s not gonna be like, “shoe” right? he will surely tell it’s a filing cabinet.