Camera flashes aren't normally the sexiest products, but Canon has made its Speedlite 470EX-AI a lot more interesting by using artificial intelligence. As photographers know, you're better off bouncing a flash off a wall or ceiling to best flatter subjects, rather than pointing it directly at them. Figuring all that out can be a pain, however, which is where Canon's "AI Bounce" tech comes in. By calculating the distance between the camera, ceiling and subject, it lets beginning photographers "utilize the ideal lighting and natural brightness of the room," Canon explains. The flash will adjust itself automatically, even if you move the camera, thanks to built-in motors at the base and hinge, and sensors in the front corner (as shown above). When you double-tap on the shutter release, the flash first points at the subject and the ceiling to calculate distances.

Enthusiastically boarding the artificial intelligence bandwagon, Ericsson will unveil a handful of "machine intelligence" products at this week's Mobile World Congress as it tries to close the gap with Chinese rival Huawei. Forthcoming products will include a chatbot designed to support mobile network technicians as well as a "self-learning" tool that shifts network resources between basestations to suit traffic levels and user needs. The basestation technology is already in trials with Vodafone Spain, which plans to use it in commercial networks this year. Pioneered by US tech giants like Google (Nasdaq: GOOG) and IBM Corp. (NYSE: IBM), artificial intelligence and machine learning are attracting growing interest from telecom network vendors hunting for new sales opportunities. Huawei Technologies Co. Ltd. plans to launch an AI-enabled computing platform called Atlas this year and has also developed a system it calls Wireless Intelligence for the automation in 5G networks of the "beamforming" process, which focuses signals on end-user devices.

The UK government has sunk £1.8 million into the development of cutting-edge artificial intelligence (AI) to bolster security and help alleviate wait times at some of the country's busiest airports. Eight projects have been chosen to trial technology that the government hopes will help bridge the gap between maintaining robust security measures and offering a quick and easy-to-use service for passengers. Security Screening Technologies, a small research team based in Derbyshire, has been given the nod to test an AI system that's been trained to identify suspicious objects in footwear, including explosives. If successful, the technology would mean passengers would no longer be required to take their shoes off during pre-flight security checks. A sophisticated scanning system is also being developed by Sequestim, a small team based in Wales, that promises to provide a way for security staff to scan passengers as they pass through security gates without the need to take off outer clothing.

We are seeing the beginning of the next big disruption in the insurance industry. Companies that don't embrace this are going to miss out on the most impactful change since the internet. In the future, people will really engage with their insurance company, and I want to tell you why by describing ways in which Artificial Intelligence (AI) will impact insurance customer experiences and fraud. I am Dave Preedy, a Digital Advisor in the UK. I am part of the Microsoft Digital Advisory Services group within Microsoft Services.

Statistical methods protecting sensitive information or the identity of the data owner have become critical to ensure privacy of individuals as well as of organizations. This paper investigates anonymization methods based on representation learning and deep neural networks, and motivated by novel information theoretical bounds. We introduce a novel training objective for simultaneously training a predictor over target variables of interest (the regular labels) while preventing an intermediate representation to be predictive of the private labels. The architecture is based on three sub-networks: one going from input to representation, one from representation to predicted regular labels, and one from representation to predicted private labels. The training procedure aims at learning representations that preserve the relevant part of the information (about regular labels) while dismissing information about the private labels which correspond to the identity of a person. We demonstrate the success of this approach for two distinct classification versus anonymization tasks (handwritten digits and sentiment analysis).

Predictive models that generalize well under distributional shift are often desirable and sometimes crucial to building robust and reliable machine learning applications. We focus on distributional shift that arises in causal inference from observational data and in unsupervised domain adaptation. We pose both of these problems as prediction under a shift in design. Popular methods for overcoming distributional shift make unrealistic assumptions such as having a well-specified model or knowing the policy that gave rise to the observed data. Other methods are hindered by their need for a pre-specified metric for comparing observations, or by poor asymptotic properties. We devise a bound on the generalization error under design shift, incorporating both representation learning and sample re-weighting. Based on the bound, we propose an algorithmic framework that does not require any of the above assumptions and which is asymptotically consistent. We empirically study the new framework using two synthetic datasets, and demonstrate its effectiveness compared to previous methods.

Abstract-- Recent successes in reinforcement learning have lead to the development of complex controllers for realworld robots.As these robots are deployed in safety-critical applications and interact with humans, it becomes critical to ensure safety in order to avoid causing harm. A first step in this direction is to test the controllers in simulation. To be able to do this, we need to capture what we mean by safety and then efficiently search the space of all behaviors to see if they are safe. In this paper, we present an active-testing framework based on Bayesian Optimization. We specify safety constraints using logic and exploit structure in the problem in order to test the system for adversarial counter examples that violate the safety specifications. These specifications are defined as complex boolean combinations of smooth functions on the trajectories and, unlike reward functions in reinforcement learning, are expressive and impose hard constraints on the system. In our framework, we exploit regularity assumptions on individual functions in form of a Gaussian Process (GP) prior. We combine these into a coherent optimization framework using problem structure. The resulting algorithm is able to provably verify complex safety specifications or alternatively find counter examples. Experimental results show that the proposed method is able to find adversarial examples quickly.

Abstract-- We propose to use neural networks for simultaneous detection and localization of multiple sound sources in human-robot interaction. In contrast to conventional signal processing techniques, neural network-based sound source localization methods require fewer strong assumptions about the environment. Previous neural network-based methods have been focusing on localizing a single sound source, which do not extend to multiple sources in terms of detection and localization. In this paper, we thus propose a likelihood-based encoding of the network output, which naturally allows the detection of an arbitrary number of sources. In addition, we investigate the use of sub-band cross-correlation information as features for better localization in sound mixtures, as well as three different network architectures based on different motivations. Experiments on real data recorded from a robot show that our proposed methods significantly outperform the popular spatial spectrum-based approaches.

Graph neural networks (GNNs) are a class of neural networks that allow to efficiently perform inference on data that is associated to a graph structure, such as, e.g., citation networks or knowledge graphs. While several variants of GNNs have been proposed, they only consider simple nonlinear activation functions in their layers, such as rectifiers or squashing functions. In this paper, we investigate the use of graph convolutional networks (GCNs) when combined with more complex activation functions, able to adapt from the training data. More specifically, we extend the recently proposed kernel activation function, a non-parametric model which can be implemented easily, can be regularized with standard $\ell_p$-norms techniques, and is smooth over its entire domain. Our experimental evaluation shows that the proposed architecture can significantly improve over its baseline, while similar improvements cannot be obtained by simply increasing the depth or size of the original GCN.

Single-cell RNA sequencing (scRNA-seq) is a fast growing approach to measure the genome-wide transcriptome of many individual cells in parallel, but results in noisy data with many dropout events. Existing methods to learn molecular signatures from bulk transcriptomic data may therefore not be adapted to scRNA-seq data, in order to automatically classify individual cells into predefined classes. We propose a new method called DropLasso to learn a molecular signature from scRNA-seq data. DropLasso extends the dropout regularisation technique, popular in neural network training, to esti- mate sparse linear models. It is well adapted to data corrupted by dropout noise, such as scRNA-seq data, and we clarify how it relates to elastic net regularisation. We provide promising results on simulated and real scRNA-seq data, suggesting that DropLasso may be better adapted than standard regularisa- tions to infer molecular signatures from scRNA-seq data.