One of the coolest hacks we saw at the TechCrunch Europe hackathon today was built by two coders from Uruguay who have been hacking their way around the world for the last few months. Truly.am lets you verify who you are talking to online. Say somebody on Facebook or a dating site sends you a picture. How do you know they really are the person in that photo? Truly.am uses some cool HTML5 tools like WebRTC and the SkyBiometry facial recognition API to verify that your conversation partners (or the people you meet on online dating sites) really look the way they say they do.
Here is how it works: somebody you don’t know sends you a picture. You then go to Truly.am, upload this picture and enter the person’s email address. That person then gets an email and has to verify his or her identity by using their webcam to take a series of images to train the recognition algorithm. Once those images are uploaded to the Truly.am servers, the facial-recognition service takes over and checks them against the original image. Once the results come back, Truly.am tells the user if the image matched and the requester, of course, also gets an email with the results.
As Agustin Haller and Dayana Jabif, the two coders behind the project, told me, you often want to remain anonymous on the net, but in some cases, you really need to know at least a bit more information about the person you are talking to. This hack was partly inspired by their own experience with Airbnb, but I have no doubt that users on online dating sites will love this product. On those sites, after all, fake photos still reign supreme.
The team plans to expand this idea to include data from LinkedIn, Xing and other professional services as well, but for now, the facial recognition app is available here and you can give it a try to verify your own potential online dates now.