The Infinity Blade Forums Have Moved

We've launched brand new Infinity Blade forums with improved features and revamped layout. We've also included a complete archive of the previous posts. Come check out the new Infinity Blade forums.
See more
See less


  • Filter
  • Time
  • Show
Clear All
new posts


    Hey all,

    Here's a project we've been working on for a while -- the core vision processing tech -- and then recently integrated with the UDK via DLL Bind:

    & go play Dungeon Defense ( )


    Great stuff!


      Looks interesting. Like natal but for any computer. The motions don't seem to match up with the persons movements very well though and seem really out of time. Could be interesting once its all figured out though, good luck with it


        it's great - will you have a development blog? I'm interested in this, couse I have a project, where I required to use 3rd party device, so I plan to integrate it via DLLBind, either. It'd be interesting for me to track your experience, if it's possible of course.


          Wow, that's really phenomenal! How are you planning on releasing it, free, paid, bundled with games, etc?


            Thanks! Yep most likely we'll release the DLL and headers free to the UDK community, though not our actual DLL source. Looking to be around June-ish for that to be ready, we want to have it nicely polished by then.

            As pointed out above, we're still tweaking it and optimizing, but that latency you still see (specifically mostly on the old laptop) will be reduced down to zero in the weeks ahead. And we're adding torso and leg capture now... soccer anyone?

            We also have an FPS demo where your movements and head orientation actually control the character's movements & looking in-game, but we're keeping a wrap on that until it's ready for unveiling.

            The key thing about the VisionPlay tech is that it works with crappy webcams and pretty low-end machines: no hardware barrier to entry. The tech is reliant on real-time computer vision body recognition research done over the last several years by one of our guy's sister companies Accelerond (, which licenses it to the military. It's pretty high-end stuff



              It seems like some of these could be easily used as a system for motion capture. Are there plans to make the source a buy-able option?


                After you've gotten through the full body capture, will you work on basic facial capture as well?