Deep Dream Vision Quest, 2016 —

A neural image synthesizer for creating multiplayer hallucinations and examining how smart objects behave in public.

Gary Boodhoo: Concept, Software, Production

Software: Python, OpenCV computer vision library, Caffe deep learning framework, bvlc_googlenet neural network, googlenet_places365 neural network, miles-deep neural network, VGG_ILSVRC_19 neural network, Linux (Ubuntu 16.04)

Hardware: Intel Core i7 6700 3.4GHz CPU, Nvidia Titan XP GPU, 16GB RAM, Small form factor micro ATX case, Mobile AV cart, 60” video display, Various USB lighting instruments, Logitech C920 webcam(s)

Although machine learning is common, its invisible. I stage science fictional experiences which exaggerate its presence in social spaces as a creative partner. My interactive video installations respond to an unpredictable world with an endless stream of recombinant imagery, which looks modern but is also ancient, and possibly universal.