Last month, Google altered our technophobic opinions on AI by using its AlphaGo program to defeat a world champion at the ancient Chinese board game of Go. Although most of us that tuned into the YouTube livestreams of the face-off were likely puzzled by the images of a computer typing out moves to a game we'd never heard of, it was a far cry from the nightmarish depictions of AI we'd become accustomed to in film and literature. Now, China's version of Google, Alibaba, is doing its bit to further familiarize us with the technology. Instead of tasking it with a board game that boasts limitless possibilities, however, it's matching it up with a more pressing (and popular) task; predicting the winner of a TV singing contest. China's popular reality TV show I'm a Singer will be getting the AI treatment, with Alibaba hoping it can outwit the public, and judges, by guessing the winner of the popular contest's finale.

In our NewsHour Shares series, we show you things that caught our eye recently on the web. Leave your suggestions in the comments below, or tweet to @NewsHour using #NewsHourShares. We might share it on air. JUDY WOODRUFF: Finally to our NewsHour Shares, something that caught our eye that we thought might be of interest to you too. Ten years ago, a car accident severed Adam Gorlitsky's spinal cord, leaving him paralyzed from the waist down.

Alphabet's intent to get rid of Boston Dynamics hasn't affected its other robotics programs, from the looks of it. On Japan's New Economic Summit stage, the Google-X-subsidiary SCHAFT unveiled a new bipedal unit that's capable of climbing stairs, carrying a loaded barbell on its "head" unit, laterally stepping through a row of seats at a soccer stadium and even maintaining balance when a section of pipe is placed under its feel. IEEE Spectrum writes that this was part of Google exec Andy Rubin's keynote at the event, but that the debut wasn't part of a product announcement or "indication of a specific product roadmap."

We revisit the development of grid based recursive approximate filtering of general Markov processes in discrete time, partially observed in conditionally Gaussian noise. The grid based filters considered rely on two types of state quantization: The \textit{Markovian} type and the \textit{marginal} type. We propose a set of novel, relaxed sufficient conditions, ensuring strong and fully characterized pathwise convergence of these filters to the respective MMSE state estimator. In particular, for marginal state quantizations, we introduce the notion of \textit{conditional regularity of stochastic kernels}, which, to the best of our knowledge, constitutes the most relaxed condition proposed, under which asymptotic optimality of the respective grid based filters is guaranteed. Further, we extend our convergence results, including filtering of bounded and continuous functionals of the state, as well as recursive approximate state prediction. For both Markovian and marginal quantizations, the whole development of the respective grid based filters relies more on linear-algebraic techniques and less on measure theoretic arguments, making the presentation considerably shorter and technically simpler.

In this paper, we design a Deep Dual-Domain ($\mathbf{D^3}$) based fast restoration model to remove artifacts of JPEG compressed images. It leverages the large learning capacity of deep networks, as well as the problem-specific expertise that was hardly incorporated in the past design of deep architectures. For the latter, we take into consideration both the prior knowledge of the JPEG compression scheme, and the successful practice of the sparsity-based dual-domain approach. We further design the One-Step Sparse Inference (1-SI) module, as an efficient and light-weighted feed-forward approximation of sparse coding. Extensive experiments verify the superiority of the proposed $D^3$ model over several state-of-the-art methods. Specifically, our best model is capable of outperforming the latest deep model for around 1 dB in PSNR, and is 30 times faster.

It's an event that plays out thousands of times across the UK every day. A consumer tries to pay for their weekly grocery shop using a credit card but with the bags packed at the till is unexpectedly told that it has been'declined'. The card is well within its credit limit, the PIN number is correct, the consumer has made numerous other purchases in the preceding weeks and yet there is no way around the reality of having no plastic money to spend. For the financial services industry these'false positives' have become a growing issue. As well as annoying customers and merchants they cost the industry in terms of the manual intervention necessary to authenticate customers and unblock cards.

Every spring, inventor Dean Kamen hosts his own sort of March Madness: a spectacle in which high-school students compete in events around the world surrounded by cheerleaders, music and entertainment. But rather than playing basketball, they're focused on building robots, an effort that culminates at the championship in St. Louis at the end of this month. Mr. Kamen, 65, is known for coming up with the Segway (the two-wheeled electric vehicle), the iBot (a stair-climbing wheelchair) and a portable dialysis machine. He considers the First Robotics Competition, now in its 25th season, one of his best ideas yet. While many young adults look up to athletes and actors as their heroes, Mr. Kamen hopes his competition--designed like a sports event, with regional brackets--will show them that there are other kinds of stars. "If you think that you can be a superstar in sports or entertainment and make the really big bucks, imagine being a superstar in tech," he says.

We provide a dependency parser for English tweets, TweeboParser . The parser is trained on a subset of a new labeled corpus for 929 tweets (12,318 tokens) drawn from the POS-tagged tweet corpus of Owoputi et al. (2013), Tweebank . These were created by Lingpeng Kong, Nathan Schneider, Swabha Swayamdipta, Archna Bhatia, Chris Dyer, and Noah A. Smith. Given a tweet, TweeboParser predicts its syntactic structure, represented by unlabeled dependencies. Since a tweet often contains more than one utterance, the output of TweeboParser will often be a multi-rooted graph over the tweet.

It is built upon one widely used machine learning package scikit-learn and two scientific computing packages Numpy and Scipy. It serves as a platform for facilitating feature selection application, research and comparative study. It is designed to share widely used feature selection algorithms developed in the feature selection research, and offer convenience for researchers and practitioners to perform empirical evaluation in developing new feature selection algorithms. Instructions of using this repository can be found in our project webpage at http://featureselection.asu.edu/

A year ago, I published my original attempt at mapping the machine intelligence ecosystem. So much has happened since. I spent the last 12 months geeking out on every company and nibble of information I can find, chatting with hundreds of academics, entrepreneurs, and investors about machine intelligence. This year, given the explosion of activity, my focus is on highlighting areas of innovation, rather than on trying to be comprehensive. Despite the noisy hype, which sometimes distracts, machine intelligence is already being used in several valuable ways.