| What's That Dog - An App For Breed Recognition |
| Written by Lucy Black | |||
| Sunday, 14 February 2016 | |||
|
Fetch! is an iOS App that recognizes dogs and classifies them by breed. A Microsoft Garage project it uses the Project Oxford APIs and can give you hours of fun. Fetch! was released just in time for major events organized by the American Kennel Club and the Westminster Kennel Club and it is a polished product that seems to be pretty successful at recognising specific breeds of dog. You simply take a photo with your smartphone, or select from the phone's scrapbook and the app will tell you what type of dog it is and what characteristics it has. Clicking on the arrow in the corner leads you to more information on the breed. If the dog’s breed is unknown, the app will show a percentage of the closest breed. Tapping the percentage rosette leads to the top five breeds that could be in the dog.
The top comment about Fetch! on the Microsoft Garage site queries why it isn't available for Windows phone: I mean... seriously, Microsoft. Fine to make an iPhone or Android app too. But do you know how frustrating it is for loyal Windows Phone users to see the big Microsoft logo beneath the Apple App store one, with nothing at all for Windows Phone. I agree with the sentiment. Back a short while Microsoft was so focused on its own OS that others simply didn't exist for it. Now what seems to matter is getting Microsoft software out there and if this means ignoring its own OS then so be it. How things change. However, for those of us who don't have an iOS device, there is the browser-based app on What-Dog.net. Like other apps that show off Project Oxford such as How-Old.net and TwinsOrNot.net this comes with a library of photos.
It also allows you to use your own photos and instantly returns a match with a breed, including its personality.
What makes it more fun is when it gets it wrong. Here's one of the IProgrammer team, Brash:
Have another go with more clues as to his size:
This dog is a terrier, a close relative of Busker, a former member of the team - and in this case the app wasn't fooled by the sunglasses!
The other fun you can have with the app is to use photos of people and discover what dogs they resemble. This gives you the chance to test the hypothesis that dog owners look like their pets, and vice versa. So using photos from Facebook, we investigated a celebrity couple - Mark Zuckerberg and his dog, Beast. Beast is a Puli, a type of Hungarian Sheepdog, but Fetch! misidentified it. Instead the result returned was Poodle for both dog and owner, with a personality trait that seems to fit:
|
Apache Lucene Improves Sparce Indexing 22/10/2024 Apache Lucene 10 has been released. The updated version adds a new IndexInput prefetch API, support for sparse indexing on doc values, and upgraded Snowball dictionaries resulting in improved tokeniza [ ... ] |
Ursina - A Game Engine Powered by Python 08/11/2024 Ursina is a new open source game engine in which you can code any type of game in Python, be it 2-D, 3-D, an application, a visualization, you name it. |
More News
|
Comments
or email your comment to: comments@i-programmer.info