By Columbia University, University of Maryland, and Smithsonian Institution
Open iTunes to buy and download apps.
Dogsnap is the second in a series of electronic field guides developed by scientists from Columbia University and the University of Maryland. This free mobile app uses visual recognition software to help identify a dog's breed from a photograph of its face.
Dogsnap is part of a larger effort to develop tools for automatic species identification to aid in the study of biodiversity. The visual recognition software Dogsnap uses to identify breeds of the domestic dog (Canis lupus familiaris) is a step toward solving the greater challenge of identifying animal species.
No upload instructions & No Camera
I feel like this app was created to give mutt owners like me a surge of hope that we could finally figure out what breed our dogs are, and then lets you down with ZERO photo upload instructions, and ZERO camera options to actually upload a picture of your puppy. Like awesome a list of breeds... I already have an app for that, it's Google. Not to mention I have a kids encyclopedia that has very vivid breed photos, so basically this app has been just as helpful as a 20 year old children's encyclopedia. Thanks Colombia... If that even is your real app developer.
I downloaded this app in hopes to find out what kind of dog my sister has. First thing that went wrong I downloaded it and then tried to open it and so I had to restart me phone. Second I go try to take a picture and there is no way to click to take the photo... I then went to the browse button and that didn't help because I already went through googles pictures of dogs that I thought looked like the one my sister has. There's just too many issues with this app no instructions to anything...
My trashcan is a Scottish Terrier???
And a picture of my wife and a picture of myself were both matched to the same breed of dog. Interesting idea, but much much more work needed. Several pictures of my mix-breed APBT were incorrectly identified as different breeds with each different shot. Seems to be due in part to the angle of the shot (accentuates the distance between the snout and eyes) as well as proximity of lens to dogs nose. Perhaps standardizing the database to a level shot at a specific distance would help?
- Category: Education
- Released: Jun 10, 2012
- Version: 1.0
- Size: 61.4 MB
- Language: English
- Seller: Peter Belhumeur
- © 2012 Columbia University, University of Maryland
Compatibility: Requires iOS 4.3 or later. Compatible with iPhone, iPad, and iPod touch.