A new collaboration between a researcher from the United States' National Security Agency (NSA) and the University of California at Berkeley offers a novel method for detecting deepfake content in a live video context – by observing the effect of monitor lighting on the appearance of the person at the other end of the video call. Popular DeepFaceLive user Druuzil Tech & Games tries out his own Christian Bale DeepFaceLab model in a live session with his followers, while lighting sources change. The system works by placing a graphic element on the user's screen that changes a narrow range of its color faster than a typical deepfake system can respond – even if, like real-time deepfake streaming implementation DeepFaceLive (pictured above), it has some capability of maintaining live color transfer, and accounting for ambient lighting. The uniform color image displayed on the monitor of the person at the other end (i.e. the potential deepfake fraudster) cycles through a limited variation of hue-changes that are designed not to activate a webcam's automatic white balance and other ad hoc illumination compensation systems, which would compromise the method. From the paper, an illustration of change in lighting conditions from the monitor in front of a user, which effectively operates as a diffuse'area light'.

The participants, from the Defence, Science and Technology Laboratory (Dstl), had three days to transform a'dumb' car into one which could navigate a course using artificial intelligence (AI) . The challenge aimed to demonstrate the benefits of reinforcement learning and the differences between the virtual and physical environments. When you are faced with a problem you have not faced before you're forced to think laterally and think outside the box. That's why we love to give our teams and particularly our early career staff these different and exciting challenges. It gives us the opportunity to engage with new technologies which are going to be really important to us in the future.

What's the news?: Deep learning is behind machine learning's most high-profile successes. But this incredible performance comes at a cost: training deep-learning models requires huge amounts of energy. Now, new research shows how scientists who use cloud platforms to train algorithms can dramatically reduce the energy they use, and therefore the emissions they create. How can they do it?: Simple changes to cloud settings are the key. Researchers created a tool that measures the electricity usage of any machine-learning program that runs on Azure, Microsoft's cloud service, during every phase of their project.

The European Parliament's Committee on Industry, Research and Energy (ITRE) recently voted in favour of Eva Maydell MEP's (BG, EPP) opinion on the Artificial Intelligence (AI) Act. The opinion received a strong endorsement from the committee with 61 votes in favour and only two against. Seven parliamentary committees are examining the proposal and Member States are still aligning their positions; but as the French pass the EU Presidency baton to the Czechs, EU40 - the platform of young pro-European MEPs - organised a multi-stakeholder discussion with leading actors and experts in the Microsoft Centre to take stock of where we are and to assess whether the EU is striking the right balance. The European Commission unveiled its proposal in April 2021 with the aim of turning Europe into a global hub for trustworthy AI. The proposal is the first of a kind.

Despite measures to limit U.S. technology exports to the Chinese military, chips designed by U.S. companies still end up in the hands of the People's Liberation Army (PLA), according to a report by the Center for Security and Emerging Technology (CSET) at Georgetown University. For the report, researchers combed through over 66,000 publicly available PLA purchase records during the eight-month period from April to November 2020 and identified 97 unique, high-end artificial intelligence (AI) chips ordered by the PLA. Nearly all of them were designed by U.S. firms Nvidia, Xilinx (now AMD), Intel, and Microsemi. The CSET report, released last month, also noted that the researchers couldn't find any public records of the PLA purchasing the high-end AI chips from any Chinese companies, including HiSilicon (Huawei), Sugon, Sunway, Hygon, and Phytium. These AI chips are critical components to the Chinese Communist Party (CCP) for the "intelligentization" (the addition of artificial intelligence to a system, according to Kaikki.org) of its military and to the regime's goal to gain dominance over the global AI design and manufacturing market.

Abstract: Multilevel linear models allow flexible statistical modelling of complex data with different levels of stratification. Identifying the most appropriate model from the large set of possible candidates is a challenging problem. In the Bayesian setting, the standard approach is a comparison of models using the model evidence or the Bayes factor. However, in all but the simplest of cases, direct computation of these quantities is impossible. Markov Chain Monte Carlo approaches are widely used, such as sequential Monte Carlo, but it is not always clear how well such techniques perform in practice.

The product segment of the AI For Cybersecurity market is bifurcated into Machine Learning,Natural Language Processing andOther. The revenue insights along with the volume forecast of each product type is incorporated in the document. Other important metrics like growth rate, market share, and other production patterns of each product type over the analysis period are given. The report categorizes the application segment of the AI For Cybersecurity market into BFSI,Government,IT & Telecom,Healthcare,Aerospace and Defense,Other,,Geographically, the detailed analysis of production, trade of the following countries is covered in Chapter 4.2, 5:,United States,Europe,China,Japan andIndia. The competitive landscape of the AI For Cybersecurity market is defined by key players such as BAE Systems,Cisco,Juniper Network,Symantec,Palo Alto Networks,Check Point,IBM,RSA Security,Fortinet andFireEye.

Artificial intelligence, and particularly machine learning (ML), is increasingly developed and deployed to support healthcare in a variety of settings. However, clinical decision support (CDS) technologies based on ML need to be portable if they are to be adopted on a broad scale. In this respect, models developed at one institution should be reusable at another. Yet there are numerous examples of portability failure, particularly due to naive application of ML models. Portability failure can lead to suboptimal care and medical errors, which ultimately could prevent the adoption of ML-based CDS in practice. One specific healthcare challenge that could benefit from enhanced portability is the prediction of 30-day readmission risk. Research to date has shown that deep learning models can be effective at modeling such risk. In this work, we investigate the practicality of model portability through a cross-site evaluation of readmission prediction models. To do so, we apply a recurrent neural network, augmented with self-attention and blended with expert features, to build readmission prediction models for two independent large scale claims datasets. We further present a novel transfer learning technique that adapts the well-known method of born-again network (BAN) training. Our experiments show that direct application of ML models trained at one institution and tested at another institution perform worse than models trained and tested at the same institution. We further show that the transfer learning approach based on the BAN produces models that are better than those trained on just a single institution's data. Notably, this improvement is consistent across both sites and occurs after a single retraining, which illustrates the potential for a cheap and general model transfer mechanism of readmission risk prediction.

The stochastic block model (SBM) is one of the most widely used generative models for network data. Many continuous-time dynamic network models are built upon the same assumption as the SBM: edges or events between all pairs of nodes are conditionally independent given the block or community memberships, which prevents them from reproducing higher-order motifs such as triangles that are commonly observed in real networks. We propose the multivariate community Hawkes (MULCH) model, an extremely flexible community-based model for continuous-time networks that introduces dependence between node pairs using structured multivariate Hawkes processes. We fit the model using a spectral clustering and likelihood-based local refinement procedure. We find that our proposed MULCH model is far more accurate than existing models both for predictive and generative tasks.

Score-based generative models exhibit state of the art performance on density estimation and generative modeling tasks. These models typically assume that the data geometry is flat, yet recent extensions have been developed to synthesize data living on Riemannian manifolds. Existing methods to accelerate sampling of diffusion models are typically not applicable in the Riemannian setting and Riemannian score-based methods have not yet been adapted to the important task of interpolation of datasets. To overcome these issues, we introduce \emph{Riemannian Diffusion Schr\"odinger Bridge}. Our proposed method generalizes Diffusion Schr\"odinger Bridge introduced in \cite{debortoli2021neurips} to the non-Euclidean setting and extends Riemannian score-based models beyond the first time reversal. We validate our proposed method on synthetic data and real Earth and climate data.