The purpose of computer vision is to realize "human-like vision" on a computer.

Existing research on non-verbal cues, e.g., eye gaze or arm movement, may not accurately present a robot's internal states such as perception results and action intent. Projecting the states directly onto a robot's operating environment has the advantages of being direct, accurate, and more salient, eliminating mental inference about the robot's intention. However, there is a lack of tools for projection mapping in robotics, compared to established motion planning libraries (e.g., MoveIt). In this paper, we detail the implementation of projection mapping to enable researchers and practitioners to push the boundaries for better interaction between robots and humans. We also provide practical documentation and code for a sample manipulation projection mapping on GitHub: https://github.com/uml-robotics/projection_mapping.

We develop a projected Wasserstein distance for the two-sample test, a fundamental problem in statistics and machine learning: given two sets of samples, to determine whether they are from the same distribution. In particular, we aim to circumvent the curse of dimensionality in Wasserstein distance: when the dimension is high, it has diminishing testing power, which is inherently due to the slow concentration property of Wasserstein metrics in the high dimension space. A key contribution is to couple optimal projection to find the low dimensional linear mapping to maximize the Wasserstein distance between projected probability distributions. We characterize the theoretical property of the finite-sample convergence rate on IPMs and present practical algorithms for computing this metric. Numerical examples validate our theoretical results.

At an after-party for the red carpet premiere of the seventh season of "Game of Thrones" at Walt Disney Concert Hall, partygoers watched as Westeros came to life on the building before them. An icy visage of the Night King loomed over the affair while dragons soared across the hall's sweeping arched facade, breathing flames that appeared to envelope the iconic structure. Santa Monica-based designer Bart Kresa created the multimedia installation for the HBO event using a technique called projection mapping. Employing high-powered video projectors and sophisticated spatial mapping software, projection mapping enables artists and designers to cast virtual graphics onto the physical world, fitting them to the exact contours and dimensions of just about any surface. Projected light shows have been popular in Europe for years, thanks to generous public funding for the arts.

You don't need special glasses to see an augmented view of the world. A new combined camera and computer can superimpose images over real-world objects without the need for a head-mounted display. Lightform connects to a video projector to beam images and animations on to surrounding objects, essentially turning any surface into a screen – a technique called projection mapping. To do this, it scans the environment using depth sensors to map the shape of objects, then tailors its lighting effects to fit. "The idea is to seamlessly merge the virtual world with the physical world, and to do it without wearing anything on your face," says Lightform CEO Brett Jones, whose firm came out of "stealth mode" this week.