A new national hub to develop artificial intelligence for challenges within healthcare will receive £250 million in Government funding. The National Artificial Intelligence (AI) Lab will bring together academics, specialists and technology companies to use technology to tackle issues such as early cancer detection, new dementia treatments and more personalised care. Tasks that will be undertaken by the AI Lab include improving cancer screening by speeding up the results of tests, including mammograms, brain scans, eye scans and heart monitoring. Predictive models will be used to predict future demand for beds, medicines, devices and surgeries. The technology will also enable the identification of which patients could be more easily treated in the community, reducing the pressure on the NHS.

The National Health Service England is planning to set up a national artificial intelligence laboratory to enhance the medical care and research facility. According to the Health Secretary, Matt Hancock said AI has'enormous power' to improve the health care facilities, and save lives. The health service has announced £250m on setting up a research lab to boost AI within the health sector. However, AI will pose new challenges in protecting patient data. Many AI tools have proven to be game-changer devices, which help doctors at spotting lung cancer, skin cancer, and more than 50 eye conditions from scans. Meanwhile, there are some tools that are yet to be used routinely across the NHS.

The company covers a good range of services based on blockchain. Last year in the month of October, the company was looking into holding an IPO (initial public offering) through stock exchange listing in Amsterdam, London or Hong Kong, as per sources. Also in the same year, the company closed an 80 million U.S. Dollar funding which was led by Korelya Capital. Market analysts and experts alike think that the company has a very potent and has the capacity to touch the value of 3 to 5 billion U.S. Dollars, the moment it gets public. This is anticipated in a coming couple of years.

Artificial intelligence may soon be deciding who lives or dies. The US Army wants to build smart missiles that will use AI to select their targets, out of reach of human oversight. The project has raised concerns that the missiles will be a form of lethal autonomous weapon – a technology many people are campaigning to ban. The US Army's project is called Cannon-Delivered Area Effects Munition (C-DAEM). Companies will bid for the contract to build the weapon, with the requirements stating it should be able to hit "moving and imprecisely located armoured targets" …

Facebook has become the latest company to admit that human contractors listened to recordings of users without their knowledge, a practice the company now says has been "paused". Citing contractors who worked on the project, Bloomberg News reported on Tuesday that the company hired people to listen to audio conversations carried out on Facebook Messenger. The practice involved users who had opted in Messenger to have their voice chats transcribed, the company said. The contractors were tasked with re-transcribing the conversations in order to gauge the accuracy of the automatic transcription tool. "Much like Apple and Google, we paused human review of audio more than a week ago," a Facebook spokesperson told the Guardian.

The fingerprints of over 1 million people, as well as facial recognition information, unencrypted usernames and passwords, and personal information of employees, was discovered on a publicly accessible database for a company used by the likes of the UK Metropolitan police, defence contractors and banks. Suprema is the security company responsible for the web-based Biostar 2 biometrics lock system that allows centralised control for access to secure facilities like warehouses or office buildings. Biostar 2 uses fingerprints and facial recognition as part of its means of identifying people attempting to gain access to buildings. Last month, Suprema announced its Biostar 2 platform was integrated into another access control system – AEOS. AEOS is used by 5,700 organisations in 83 countries, including governments, banks and the UK Metropolitan police.

In this Thursday, July 11, 2019, photograph, United States Department of Agriculture intern Alex Olsen prepares to place down a drone at a research farm northeast of Greeley, Colo. After a brief, snaking flight above the field, the drone landed and the researchers removed a handful of memory cards. Back at their computers, they analyzed the images for signs the corn was stressed from a lack of water. This U.S. Department of Agriculture station outside Greeley and other sites across the Southwest are experimenting with drones, specialized cameras and other technology to squeeze the most out of every drop of water in the Colorado River – a vital but beleaguered waterway that serves an estimated 40 million people. Should they still be able to use it?

Deep learning has penetrated all aspects of our lives and brought us great convenience. However, the process of building a high-quality deep learning system for a specific task is not only time-consuming but also requires lots of resources and relies on human expertise, which hinders the development of deep learning in both industry and academia. To alleviate this problem, a growing number of research projects focus on automated machine learning (AutoML). In this paper, we provide a comprehensive and up-to-date study on the state-of-the-art AutoML. First, we introduce the AutoML techniques in details according to the machine learning pipeline. Then we summarize existing Neural Architecture Search (NAS) research, which is one of the most popular topics in AutoML. We also compare the models generated by NAS algorithms with those human-designed models. Finally, we present several open problems for future research.

We study the robust one-bit compressed sensing problem whose goal is to design an algorithm that faithfully recovers any sparse target vector $\theta_0\in\mathbb{R}^d$ uniformly $m$ quantized noisy measurements. Under the assumption that the measurements are sub-Gaussian random vectors, to recover any $k$-sparse $\theta_0$ ($k\ll d$) uniformly up to an error $\varepsilon$ with high probability, the best known computationally tractable algorithm requires $m\geq\tilde{\mathcal{O}}(k\log d/\varepsilon^4)$ measurements. In this paper, we consider a new framework for the one-bit sensing problem where the sparsity is implicitly enforced via mapping a low dimensional representation $x_0 \in \mathbb{R}^k$ through a known $n$-layer ReLU generative network $G:\mathbb{R}^k\rightarrow\mathbb{R}^d$. Such a framework poses low-dimensional priors on $\theta_0$ without a known basis. We propose to recover the target $G(x_0)$ via an unconstrained empirical risk minimization (ERM) problem under a much weaker sub-exponential measurement assumption. For such a problem, we establish a joint statistical and computational analysis. In particular, we prove that the ERM estimator in this new framework achieves an improved statistical rate of $m=\tilde{\mathcal{O}}(kn \log d /\varepsilon^2)$ recovering any $G(x_0)$ uniformly up to an error $\varepsilon$. Moreover, from the lens of computation, despite non-convexity, we prove that the objective of our ERM problem has no spurious stationary point, that is, any stationary point are equally good for recovering the true target up to scaling with a certain accuracy. Furthermore, our analysis also shed lights on the possibility of inverting a deep generative model under partial and quantized measurements, complementing the recent success of using deep generative models for inverse problems.

Multimodal learning aims to discover the relationship between multiple modalities. It has become an important research topic due to extensive multimodal applications such as cross-modal retrieval. This paper attempts to address the modality heterogeneity problem based on Gaussian process latent variable models (GPLVMs) to represent multimodal data in a common space. Previous multimodal GPLVM extensions generally adopt individual learning schemes on latent representations and kernel hyperparameters, which ignore their intrinsic relationship. To exploit strong complementarity among different modalities and GPLVM components, we develop a novel learning scheme called Harmonization, where latent model parameters are jointly learned from each other. Beyond the correlation fitting or intra-modal structure preservation paradigms widely used in existing studies, the harmonization is derived in a model-driven manner to encourage the agreement between modality-specific GP kernels and the similarity of latent representations. We present a range of multimodal learning models by incorporating the harmonization mechanism into several representative GPLVM-based approaches. Experimental results on four benchmark datasets show that the proposed models outperform the strong baselines for cross-modal retrieval tasks, and that the harmonized multimodal learning method is superior in discovering semantically consistent latent representation.