Neural Information Processing Systems http://nips.cc/

Color constancy is the recovery of true surface color from observed color, and requires estimating the chromaticity of scene illumination to correct for the bias it induces. In this paper, we show that the per-pixel color statistics of natural scenes--without any spatial or semantic context--can by themselves be a powerful cue for color constancy. Specifically, we describe an illuminant estimation method that is built around a "classifier" for identifying the true chromaticity of a pixel given its luminance (absolute brightness across color channels). During inference, each pixel's observed color restricts its true chromaticity to those values that can be explained by one of a candidate set of illuminants, and applying the classifier over these values yields a distribution over the corresponding illuminants. A global estimate for the scene illuminant is computed through a simple aggregation of these distributions across all pixels.

In a world that remains anything but normal, gaming has become a source of companionship and catharsis. The game industry has struggled to keep up as major releases like Horizon Forbidden West and God of War Ragnarök have once again been delayed by the pandemic. But you'd be hard pressed to find a year with more breadth. Over 11 thousand unique titles were released to the online marketplace Steam, thousands more than any other year. Consoles like the Xbox Series X, PS5, and Nintendo Switch each deepened their catalogs with new franchises and remasters of old classics, and independent games enjoyed heightened visibility at The Game Awards and on platforms like YouTube and Twitch.

For the most part, a lot of the things that made me love the franchise in the first place are back. I love the small-town setting. I love how you really get to know the place just as well as the characters. The characters themselves are really strong. The fact that there's a small cast that recur throughout all five chapters instead of just sporadically appearing like they did in "Life is Strange 2″ helps you connect with them more.

Most glaringly, "True Colors" centers Asian American characters, though its story could be about anyone of any background. The game never talks about race, except for showing an Asian-style shrine that Alex puts up to commemorate Gabe's life and putting a few rude comments on the game's version of Facebook or Nextdoor, MyBlock. And so, its diverse casting feel inconsequential. We don't live in a post-racial society -- as nice as that would be -- and by avoiding the subject almost entirely, "True Colors" draws in fans who want representation without living up to those fans' excitement and expectations.

A remastered collection of the first two Life is Strange games will arrive a bit later than expected. Square Enix says it's pushing back the bundle of Life Is Strange and Life Is Strange: Before the Storm to early 2022. The Life Is Strange Remastered Collection was originally scheduled for September 30th. "Due to the ongoing challenges of the worldwide pandemic, we want to alleviate any additional pressure on the Life is Strange team by giving more time between the release of Life is Strange: True Colors and the Life is Strange Remastered Collection," Square Enix wrote in a tweet. While the delay might be disappointing to some, it's good to see the developers getting more breathing room. The new versions will include "vastly improved character animation" drawn from full facial motion capture.

In "True Colors," players control a new protagonist, a young Asian American woman named Alex Chen, played by Erika Mori, who moves to the town of Haven Springs. Upon her arrival, she's reunited with her brother Gabe for a short time, until a mysterious tragedy occurs -- or as Square Enix describes it, a "so-called accident." To find out what happened, along with uncovering "dark secrets" within the town, Alex uses her psychic power of Empathy. This allows her to absorb and manipulate the emotions of others. At the beginning of the game, Alex refers to her power as a curse, as this power can take over her own emotions if she's not careful.

Underwater photography is hard to get right. Special filters, artificial lights, and top-of-the-line underwater cameras can help, but there's still a lot of water between the camera and the object in the photo. We've become accustomed to the blue-green tint of underwater photography. How would the ocean look without water? What are the true colors of a coral reef?

Color constancy is the recovery of true surface color from observed color, and requires estimating the chromaticity of scene illumination to correct for the bias it induces. In this paper, we show that the per-pixel color statistics of natural scenes---without any spatial or semantic context---can by themselves be a powerful cue for color constancy. Specifically, we describe an illuminant estimation method that is built around a classifier for identifying the true chromaticity of a pixel given its luminance (absolute brightness across color channels). During inference, each pixel's observed color restricts its true chromaticity to those values that can be explained by one of a candidate set of illuminants, and applying the classifier over these values yields a distribution over the corresponding illuminants. A global estimate for the scene illuminant is computed through a simple aggregation of these distributions across all pixels. We begin by simply defining the luminance-to-chromaticity classifier by computing empirical histograms over discretized chromaticity and luminance values from a training set of natural images. These histograms reflect a preference for hues corresponding to smooth reflectance functions, and for achromatic colors in brighter pixels. Despite its simplicity, the resulting estimation algorithm outperforms current state-of-the-art color constancy methods. Next, we propose a method to learn the luminance-to-chromaticity classifier end-to-end. Using stochastic gradient descent, we set chromaticity-luminance likelihoods to minimize errors in the final scene illuminant estimates on a training set. This leads to further improvements in accuracy, most significantly in the tail of the error distribution.