Artificial Itelligence can recognise breeds of dogs
Ml5.js is a library that gives access to machine learning algorithms and models in the browser.
In this library there are codes, examples, tutorials and sample data.
There are different ways to use it:
Artificial Intelligence is able to recognise images.
Computer can register many images and then can recognise the thing on the image thanks to its database.
It is based on similarity between the images in the database and the images we give to the Artificial Intelligence.
These videos explain how it works:
MobileNet tells you what the AI recognizes and also the pourcent of confidence.
First we need to feed the Artificial Intelligence with data. We give classifiations to the machine. For instance we give different breeds of dogs, consequently the application will recognise the breeds of the dogs in the images.
Dog breeds are usually categorized by the functional type from which the breed was developed. The basic types are companion dogs, guard dogs, hunting dogs, herding dogs and working dogs. Breeds listed here may be traditional breeds with long histories as registered breeds, rare breeds with their own registries, or new breeds that may still be under development.
For example there are the Golden Retriver, the Bulldog and the Husky.
To be sure of the breed of one dog, you should check with a veteritarian.
The more images the Artificial Intelligence encounters, the more reliable it is.
You can find many images and manys breeds of dogs on the Internet.
Once you have find an image of dog, you can bring it in the application.
Then the application looks after in its data and recognises the breed of the dog.
That's it ! ;)
This application is very nice to use and can be useful to discover new breeds of dogs.
