What's the point?
In attempting to promote this kickstarter to various groups such as reddit programming and pets, I keep coming across the same tired objections.
The money question, now has an answer for it.
i would now like to address the next objection- What's the point?
The objections go something like this:
What's the point? It seems very impractical. What if my cat gets lost?
To the above questions. Thank you for taking an interest in this kickstarter.
However, the real vision behind this project is somehow being lost.
Authentication is merely the simplest implementation that I could think of to build for what I am really attempting to accomplish.
In 2004 DARPA issued a grand challenge. Read about it here: http://www.darpa.mil/grandchallenge04/ and here: http://en.wikipedia.org/wiki/DARPA_Grand_Challenge
In this challenge, they put out a sizable prize to the team which could build a fully autonomous vehicle which could navigate 150 miles of desert terrain. They all failed in 2004. The following year in 2005 Stanford University took first place and crossed the finish line. Four other teams completed the race.
I just want to make sure everyone know what autonomous means ( this isn't to talk down to anyone)
That's right. Stanford and four other universities built a car that can drive itself!
Pretty amazing, right?
The Computer Vision (OpenCV) component of this winning vehicle is the same Computer Vision engine I am proposing to use to build cat face recognition into.
See, Authentication is a game of true and false. You either gain access or you don't.
By building an Open Source statistical classifier which can detect and recognize cats, all sorts of projects become possible. Anything you can hook up an arduino to, anything which interacts with cats, etc can become smarter.
Imagine being able to hook an arduino up to a coffee pot and make it bark anytime a cat jumps on the counter and comes near the coffee pot. There are hundreds of examples of where having the ability to detect a cat could be incredibly helpful for a hobbyist project.
Again, with an open source license, car manufacturer could build this into newer model vehicles and alert a driver when a cat is on the road. The possibilities are endless.
This is why being cross platform is so important.
I not only would be releasing the classifier, but the data set I used in the first place.
I realize this is a long post.
However, I encourage you to allow your imagination to run wild. The possibilities are incredible.
- First created · 0 backed
- Has not connected Facebook
Pledge $1 or moreYou selected
Thank you. Every dollar counts. Your name on our website as a supporter.
Pledge $5 or moreYou selected
access to our blog which will post updates to the project status as well as a special thank you listing on the blog.
Pledge $10 or moreYou selected
You get to vote on the default access granted and access denied sounds (cat noises)
Pledge $15 or moreYou selected
Your name and website link in the credits on this cat face recognition authentication program.
Pledge $25 or moreYou selected
A signed postcard of Emmy Lou the kitty. This post card will be signed by the full team. Kitty print included.
Pledge $50 or moreYou selected
A signed DVD of the complete source code and compiled copies for Mac, Windows, and Linux. Installers for each mentioned operating systems provided . 64 bit and 32 bit binaries will be provided.
Pledge $100 or moreYou selected
Early access to the source code. Plus a limited edition 'i can haz login' t-shirt in your size. For international orders, please add $10USD
Pledge $250 or moreYou selected
Pick the codename for the release number of your choice as available and name a function.
Pledge $500 or moreYou selected
Meet the Development team over Skype and learn either the basics of OpenCV or PAM ( Pluggable Authentication Module) from Andrew McElroy. Expected Course time 6 hours.
Pledge $1,000 or moreYou selected
Get into the mind of the development team! For up to 20 hours total we will either code review with you line by line on this project so that you know exactly whats going on, or we will teach PAM or OpenCV at an intermediate to advanced level. You can expect to either be able to administrate and write modules for PAM or be able to fully use OpenCV (2D vision mode not 3D) as well as create your own statistical classifiers. (We will not go over each and every existing classifier in OpenCV as there are several hundred already. A working knowledge of statistics and calculus is expected for OpenCV training)
- (60 days)