How would neural network detect this object? Is it a cat or a bread?
How would neural network detect this object? Is it a cat or a bread?
Kayden Carter
Other urls found in this thread:
Tyler Price
bread doesn't have eyes, ears, or a mouth what do you think
Luke Wilson
you can do multi label classification. it would be both bread and a cat
Evan Wright
but cats don't have bread decoration on top of them.
Jaxon Murphy
It would be classified as a anomaly , and Terminator bot commissioned to exterminate it.
Andrew Morris
Depends on what you train it to recognize. If you train it to detect cats in bread loaves (will require many examples) theb it will even detect cats in bread loaves no problem.
It all depends on the weights. The same neural network architecture with different weights will recognize different objects (see famous architectures such as ResNet50, VGG, LeNet etc)
Jose Perry
>submit 4000 Google Captcha
>2000 "Click on all catsats, if you see none click skip"
>2000 "click on all breads, if you none click skip"
>Evaluate results and determine if it's a cat or a bread
Hunter Barnes
Bread, most likely
Daniel Gonzalez
Depends on your training examples really. NNs aren't really smart, they just find certain patterns in pixels. If you had many breads in your dataset that looked kind of like this pic, it might say bread, otherwise it might say cat. It might say bread because the cats you had in your data were slim, while if you had many fat cats it might say cat.
Carson Harris
epic post regditbro xDDDDD le robot apocalypse is le soon!!!
Jaxon Ward
Not only are there numerous types of NNs, such as Convolutional NN and Gradient Decent (Descent? cba) NN, but their training data also matters. I believe the latter is the most significant factor dictating the ultimate output of a NN. Has it come in contact with felines or baked goods?
Lucas Phillips
Do you ever wonder if there's really an AI we're training with the capchas ?
Seeing the size of the userbase, we might just be solving that stuff in real time and we'll never know
Lucas Diaz
>bread decoration
10/10
Leo Adams
That is bread.
Robert Evans
Google offers their ML shit for free and you still need somebody to hold your hand.
Thomas Bennett
why does that kitty look like bread??
Landon Howard
NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO I DON'T WANT TO BE BREAD AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
Dylan Barnes
How do i use it for free
they want a credit card to use the API wtf
Robert Rodriguez
>cuisine
Oh China, you so crazy.
Gavin Davis
Is it a bike or a bird? Or a bird on a bike??
Congratulations OP you just showed everyone here what the difference between biological neurons and artificial ones are.
Samuel Ortiz
link plz
Anthony Morgan
Caleb Jenkins
The bread classifier wikl be X % certain that it’s a bread, the cat classifier will be Y % certain that it’s a cat, whichever is higher wins. It mainly depends on he dataset it was trained on.
Connor Nelson
damn imagine typing like this unironically lol
Grayson Rivera
>Cat 100%
>Bread 100%
>Cursed 100%
Kevin Hernandez
Cnns have been shown to overvalue texture over shape. It would likely choose bread as the class unless it's been properly regularized.
Ryan Gutierrez
This cat clearly does
Nolan Clark
>Small to Medium-sized cats
That's very kind of them.