Laossj - 无标题

SP. Augmented Reality. Psycho-Pass (2012)

4DAR ARKit Unity

Developer  应高选 has been experimenting with 4DViews’ free 4D captures and shares the results - particularly striking is this one using the new Apple ARKit and Unity software:

4DAR with ARKit and Unity3D, real man and real scale. iPhone6s test.

Here is an example using the same assets at a smaller scale:

应高选’s YouTube channel can be found here

4DViews on PK (from last week) Here

[Update 10/07/17]

Here is a video from 4DAR demonstrating how to put together your ownin 5 minutes:

Almost realtime visual tutorial on using 4DViews volumetric capture sequence with Unity and Apple ARKit, for fast hologram display

Mario Kart 64 with Neural Evolution of Augmenting Topologies (NEAT)

Project from NIck Nelson applies Neural Network learning to playing Mario Kart 64 with successful results:

This is NOT a human playing the game, it is in fact the program I wrote. It is a special kind of machine learning that models biological evolution to evolve “species” to find the optimal solution to the problem. In this case the problem is Mario Kart 64! This run is the result of about two days of training. 

Code for the project can be found here

Channel your inner #WonderWoman and discover your coding powers! https://goo.gl/n0TMGq

Adversarial Examples that Fool Detectors

Computer Vision research from Jiajun Lu, Hussein Sibai and Evan Fabry examines blocking neural network object detection using what appears to look like DeepDream-esque camouflage:

An adversarial example is an example that has been adjusted to produce a wrong label when presented to a system at test time. To date, adversarial example constructions have been demonstrated for classifiers, but not for detectors. If adversarial examples that could fool a detector exist, they could be used to (for example) maliciously create security hazards on roads populated with smart vehicles. In this paper, we demonstrate a construction that successfully fools two standard detectors, Faster RCNN and YOLO. The existence of such examples is surprising, as attacking a classifier is very different from attacking a detector, and that the structure of detectors - which must search for their own bounding box, and which cannot estimate that box very accurately - makes it quite likely that adversarial patterns are strongly disrupted. We show that our construction produces adversarial examples that generalize well across sequences digitally, even though large perturbations are needed. We also show that our construction yields physical objects that are adversarial. 

The paper can be found here

holy motors_2017

5 Mysterious Posts Found On Reddit That STILL Remain Unexplained…

This is actually the 2nd Machine learning program!

Plastic Girls

Short film by Nils Clauss explores the spaces that use female-form robotic mannequins around South Korea to attract business:

With slow graceful movements and digitally generated messages, Plastic Girls make a conspicuous contribution to the sexualisation of public space in front of their owners premises.

PLASTIC GIRLS is the last film of a Korea related trilogy following BIKINI WORDS and LAST LETTERS:

vimeo.com/151000182 vimeo.com/195284427

Link

Piano player wears an eye tracker so you can see exactly where their eyes move to as they play. Amazing video.