Over the last few years, computer vision and motion capture technology has really advanced at an incredible rate. Things that seemed like science fiction not too long ago can be purchased for very little and in your home today. Things like the Kinect and now the Leap Motion controller are getting people to rethink even the way they interact with computers. And this space is really still in its earliest days. So for this week's awesome stuff
post (in which we highlight interesting crowdfunding projects we come across) we'll highlight two new projects that deal with computer vision and motion capture -- both of which work with merely a webcam, rather than any fancy hardware.
- First up is the FaceRig, which uses your everyday webcam to not only capture your facial expressions, but translate them into the faces of characters in an insanely realistic manner. Full disclosure: after seeing this video I plunked down some money. I have no idea what I'd use it for, but it's super cheap and it just looks awesome. It really wasn't that long ago that filmmakers who wanted to do motion capture work to create a computer generated character would have to do an astounding amount of work to match the characters to actors' actual facial expressions. And yet, here's a system that appears to be able to do it on the fly in a manner that looks incredibly good, at least from the demo they show.
Not only does the product look kind of amazing, they're pricing it to be incredibly affordable. Getting a license via the crowdfunding is just $5. They're shooting for $120,000 which is pretty ambitious, especially for an IndieGogo project, but within just a few days they already shot past $50,000 and with a month and a half to go, it seems likely that they'll fly past their goal and end up with significantly more than that. Combination of amazing product plus cheap prices will do that.
- Next up, we've got the VMX Project, which is more of a platform for developers to be able to add computer vision to anything. They're offering it as a service, so rather than having to install software, developers can just use their API and add in computer vision features via (what they claim) is a super simple user interface. The concept of making computer vision more widespread and usable by developers, without each developer having to reinvent the wheel, or hire experts in computer vision, seems like a super powerful idea in driving this area forward.
These guys have a tougher job ahead of them. Even as the product looks cool, it's very much focused on developers, rather than the general consumer market, so there's a significantly smaller audience. These guys are targeting $100,000, which is a pretty big ask, all things considered, and the project hasn't received very much at all yet -- so it seems like a good chance it won't hit that goal.
That's it for this week. Go interact with something.