Researchers at the University of Washington have produced a photo-realistic former US President Barack Obama. They have used Artificial intelligence (AI) to precisely mirror how Barack Obama moves his mouth when he speaks. This technique allows them to put any words into his mouth and create an unbelievably convincing public announcement. The reason for Obama being the chosen candidate for this work is down to the countless hours of high-definition video of him available across the web. The research team had an artificial neural network analyse millions of frames of video to note how Obama's facial features moved as he was talking; putting most of the focus on his mouth.

The White House on Thursday hosted a summit on artificial intelligence, bringing together more than 100 business leaders, government officials and technical experts to discuss AI's potential for making sense of the data that is inundating healthcare and other industries. Representatives from agriculture, energy, financial services, healthcare, manufacturing and transportation attendaed the summit to discuss a number of topics, including research and development, workforce development, regulatory barriers to AI innovation and sector-specific applications of the technology. Tech giants such as Amazon, Facebook, Google and Microsoft also participated. "Artificial intelligence holds tremendous potential as a tool to empower the American worker, drive growth in American industry and improve the lives of the American people," said Michael Kratsios, deputy assistant to the President for technology policy. "Our free market approach to scientific discovery harnesses the combined strengths of government, industry and academia, and uniquely positions us to leverage this technology for the betterment of our great nation."

Neural Machine Translation (NMT) performs poor on the low-resource language pair $(X,Z)$, especially when $Z$ is a rare language. By introducing another rich language $Y$, we propose a novel triangular training architecture (TA-NMT) to leverage bilingual data $(Y,Z)$ (may be small) and $(X,Y)$ (can be rich) to improve the translation performance of low-resource pairs. In this triangular architecture, $Z$ is taken as the intermediate latent variable, and translation models of $Z$ are jointly optimized with a unified bidirectional EM algorithm under the goal of maximizing the translation likelihood of $(X,Y)$. Empirical results demonstrate that our method significantly improves the translation quality of rare languages on MultiUN and IWSLT2012 datasets, and achieves even better performance combining back-translation methods.

Single-trial classification of event-related potentials in electroencephalogram (EEG) signals is a very important paradigm of brain-computer interface (BCI). Because of individual differences, usually some subject-specific calibration data are required to tailor the classifier for each subject. Transfer learning has been extensively used to reduce such calibration data requirement, by making use of auxiliary data from similar/relevant subjects/tasks. However, all previous research assumes that all auxiliary data have been labeled. This paper considers a more general scenario, in which part of the auxiliary data could be unlabeled. We propose active semi-supervised transfer learning (ASTL) for offline BCI calibration, which integrates active learning, semi-supervised learning, and transfer learning. Using a visual evoked potential oddball task and three different EEG headsets, we demonstrate that ASTL can achieve consistently good performance across subjects and headsets, and it outperforms some state-of-the-art approaches in the literature.

Ensemble learning is a powerful approach to construct a strong learner from multiple base learners. The most popular way to aggregate an ensemble of classifiers is majority voting, which assigns a sample to the class that most base classifiers vote for. However, improved performance can be obtained by assigning weights to the base classifiers according to their accuracy. This paper proposes an agreement rate initialized maximum likelihood estimator (ARIMLE) to optimally fuse the base classifiers. ARIMLE first uses a simplified agreement rate method to estimate the classification accuracy of each base classifier from the unlabeled samples, then employs the accuracies to initialize a maximum likelihood estimator (MLE), and finally uses the expectation-maximization algorithm to refine the MLE. Extensive experiments on visually evoked potential classification in a brain-computer interface application show that ARIMLE outperforms majority voting, and also achieves better or comparable performance with several other state-of-the-art classifier combination approaches.

There are many important regression problems in real-world brain-computer interface (BCI) applications, e.g., driver drowsiness estimation from EEG signals. This paper considers offline analysis: given a pool of unlabeled EEG epochs recorded during driving, how do we optimally select a small number of them to label so that an accurate regression model can be built from them to label the rest? Active learning is a promising solution to this problem, but interestingly, to our best knowledge, it has not been used for regression problems in BCI so far. This paper proposes a novel enhanced batch-mode active learning (EBMAL) approach for regression, which improves upon a baseline active learning algorithm by increasing the reliability, representativeness and diversity of the selected samples to achieve better regression performance. We validate its effectiveness using driver drowsiness estimation from EEG signals. However, EBMAL is a general approach that can also be applied to many other offline regression problems beyond BCI.

To choose a suitable multi-winner rule, i.e., a voting rule for selecting a subset of k alternatives based on a collection of preferences, is a hard and ambiguous task. Depending on the context, it varies widely what constitutes the choice of an "optimal" subset. In this paper, we offer a new perspective to measure the quality of such subsets and--consequently-- multi-winner rules. We provide a quantitative analysis using methods from the theory of approximation algorithms and estimate how well multi-winner rules approximate two extreme objectives: diversity as captured by the (Approval) Chamberlin-Courant rule and individual excellence as captured by Multi-winner Approval Voting. With both theoretical and experimental methods we classify multi-winner rules in terms of their quantitative alignment with these two opposing objectives.

Maven is also only one of hundreds of AI initiatives being pursued across the Pentagon. So many programs spread out over such a large organization can be the stuff of nightmares for military planners, but the building's hard-charging new undersecretary for research and engineering, Michael Griffin, said recently that he is setting up a Joint Artificial Intelligence Center that will will tie together the military's efforts with those of the Intelligence Community, allowing them to combine efforts in a breakneck push to move government's AI initiatives forward.

What is Robotic Process Automation really all about? I'm Shaun Williamson, Senior Product Owner and I'm part of a team exploring how robotics and intelligent automation can be exploited in the DWP Digital environment. Although'the robots' are not really going to take over the world, this stuff is big business, and a growing technology trend. If you are not a'robotista' already, then it's time to find out more. We have a team of colleagues learning more about the solutions behind Artificial Intelligence and how they can be applied to DWP's business environment for greatest value and return on investment.

With the rise of the autonomous machine revolution, there has been an acceleration in the development and deployment of robotics across a broad range of industries. This includes fields such as emergency response where each minute is critical as victims must wait for help or supplies to arrive. Yet high-tech drones could decrease that waiting time significantly. SURVICE Engineering company is contributing to these efforts by producing droids that are capable of speedily delivering medical supplies and supporting the Department of Defense.