Incredible progress has been made in object recognition over the past decade. New algorithms are able to detect generic objects, such as people, cars, trucks, animals, and other common objects, with high accuracy. In this talk, we describe how a combination of optimizations on CUDA-enabled GPUs, and novel improvements in computer vision algorithms enabled us to accelerate object recognition by two orders of magnitude, enabling for the first time, real-time state-of-the-art object recognition of dozens of object types.
Slides can also be found at:
I first saw these guys at the NVIDIA GPU Technology Conference, then again at Jen-Hsen's RTT Excite keynote.
It's absolutely awesome and it's technology like this that will literally change everything (again) when it comes to 3D graphics. Having fast powerful computer vision and the ability to accurately recognize key objects, like houses, people, animals, seat, cars, enables a wealth of new App experiences, especially when it comes to Augmented Reality.