Yes, it's that time again. SIGGRAPH 2014 is only a short time away and there is a showcase video that lets you see some of the amazing things coming your way.
SIGGRAPH is not only for graphics researchers; it is a view of what we can expect in games, future UIs and apps. For a round up of some of the things you can expect to see at this year's SIGGRAPH, take a look at the video:
What's the best or most interesting?
A very difficult question. But one thing I'm sure of - the voice over is terrible. In fact if you read the titles of the papers (they are in the box at the bottom left) and listen to the voice over at the same time it produces a strange "other worldy" effect. The academic tone of "Discrete Stochastic Microfacet Models" with the overlay of "shiny round things" said in a slightly manic DJ style is bizarre.
As to the papers - well all the "shiny round things" papers look real fun. But for me the one highlight (pun intended) is "Learning Bicycle Stunts", the next step for which is to use a real robot. I'd also draw attention to the boring sounding "Inverse-Foley Animation: Synchronizing Rigid-Body Motions to Sound", which has lots of potential and is far from boring. "3D Object Manipulation In a Single Photograph using Stock 3D Models" is very impressive and could give rise a whole new type of app. The same could be said for "High-Contrast Computational Caustic Design" - I've seen one off clever images created by caustic curve reflection but the idea that you can automate the process using a 3D printer never occurred to me. Finally what's not to love about melting bunnies, covered in Augmented MPM for Phase-Change and Varied Materials.
My "whoa" why didn't I think of that prize goes to "Pixie Dust". The idea of setting up a sound field to levitate and control small particles is cool but I have to admit to doubting that it has any real practical application.
And finally I really want to know how to create a depth camera with an old webcam plus a few bits of sticky plastic.
If you have been reading our reports on adversarial images, the headline should come as no surprise. What is a surprise is the way that AI researchers are regarding such images as security threats rat [ ... ]