In Your Face! Huiiz Offers An Innovative Facial Recognition Search Engine
A picture is worth 1000 words, and these days, can be seen by thousands of people – whether you know it or not. With Huiiz (pronounced: hue-eye-eez), an innovative facial recognition search engine (think Google, but for your face), you are given the only tool you need to sniff-out the deepest corners of the web to see where your pictures have landed.
A friend of mine commented to me the other day that she was informed by “girlfriend revenge site” that explicit photos of her were posted on their page. The site stated that they would remove the photos for $400 USD – an extortive measure many of these cesspools use to muscle unsuspecting women whose privacy has been invaded. Many of these sites use a photo pulled from one of the victim’s social media pages as a way to “identify” them among the risqué pictures.
Using Google images and searching your name can only yield certain results – ones that are tagged with corresponding keywords. Places like the aforementioned “revenge” sites, strip those keywords away and can easily pull your images for their malicious intents.
It turned out, the photos ended up not being photos of my friend. However, an innovative facial search engine like Huiiz could’ve been used as countermeasure to unauthorized uses of your pictures. Something that happens way too often these days.
How it works.
Search engines are meant to be simple and user-friendly, and Huiiz’s intuitive, “drag-and-drop” design allows you search with ease.
If a picture is too large, the engine crops it automatically; if it contains more than one person, an editing prompt allows you to select the correct face to be searched. Once the image is inputted Huiiz’s algorithm prowls the web, and according to the creators, has a 93% recognition rate. What is interesting about the technology behind the engine, is that it “learns” and evolves with addition of more users. The bigger the user base, the better the results.
There is always the chance of not getting it completely right – as we humans live among doppelgangers and look-alikes that share facial features which can fool even the smartest algorithm. However, the creators claim that their code is “able to avoid wasting computer resources in areas of the web that are not important,” and has identified and indexed “tens of millions” of pictures already.
A voting system will also be implemented to allow beta testers to ability to provide feedback on the features they desire.
Currently, Huiiz is in its infancy, utilizing a Kickstarter campaign to build hype and potential backers in order to finish its three main components: the website, mobile app, and webcrawler infrastructure. Backers can easily join the party with a $10 donation and can pay upwards of $1250 for a “developer’s package,” allowing you full access to the engine’s API for your own personal projects.
Creator and team lead, Sebastien Nadeu, has been working professionally with images for over 10 years and started this project in late 2012. Still, with the recent backlash from Oculus Rift supporters on Kickstarter, many are wary that a big company like Facebook could easily swoop in and pluck the campaign – one that would make a lot of sense the social media giant.
You can find all the information about Huiiz and their innovative facial recognition search engine here.