On October 10, 2018, Huawei officially released a full-stack all-scenario AI solution. 'Full-stack' means the solution provides powerful and cost-effective computing power, as well as low-barrier application development platforms that make AI data modeling, model training, and application development simpler, more agile, and more efficient. As an'all-scenario' solution, it is able to drive pervasive intelligence across various device, edge, and cloud scenarios. Ascend 310 is the industry's most powerful AI SoC chip for edge computing scenarios. Currently, the most typical edge computing scenarios are security and protection, autonomous driving, and smart manufacturing.

Ask someone to list life's essentials without which they can't survive, and chances are that their phone will feature near the top. With the increasing use of smartphones, the telecommunication & media industry has become an increasingly critical player in our lives, arguably representing the backbone of digital disruption. Yet service providers still have much to learn. Imagine driving to work and your navigation app prompts you to change routes to avoid a traffic buildup. The source of this information is a combination of artificial intelligence (AI) and machine learning (ML).

At trade fairs like the Mobile World Congress and IFA, Huawei, Samsung and Qualcomm are the main companies that focus on AI. You can also throw in Apple's new bionic processor and Google's Pixel Visual Core into that mix. Of course, this is only the current state of AI functions. AI is growing rapidly in the world of smartphones, and in the future we'll see many more applications. Development is progressing and we're still just at the beginning. Yet, even today there are exciting AI functions you can enjoy, if you have the right smartphone.

The digital telecommunications receiver is an important context for inference methodology, the key objective being to minimize the expected loss function in recovering the transmitted information. For that criterion, the optimal decision is the Bayesian minimum-risk estimator. However, the computational load of the Bayesian estimator is often prohibitive and, hence, efficient computational schemes are required. The design of novel schemes, striking new balances between accuracy and computational load, is the primary concern of this thesis. Two popular techniques, one exact and one approximate, will be studied. The exact scheme is a recursive one, namely the generalized distributive law (GDL), whose purpose is to distribute all operators across the conditionally independent (CI) factors of the joint model, so as to reduce the total number of operators required. In a novel theorem derived in this thesis, GDL, if applicable, will be shown to guarantee such a reduction in all cases. An associated lemma also quantifies this reduction. For practical use, two novel algorithms, namely the no-longer-needed (NLN) algorithm and the generalized form of the Markovian Forward-Backward (FB) algorithm, recursively factorizes and computes the CI factors of an arbitrary model, respectively. The approximate scheme is an iterative one, namely the Variational Bayes (VB) approximation, whose purpose is to find the independent (i.e. zero-order Markov) model closest to the true joint model in the minimum Kullback-Leibler divergence (KLD) sense. Despite being computationally efficient, this naive mean field approximation confers only modest performance for highly correlated models. A novel approximation, namely Transformed Variational Bayes (TVB), will be designed in the thesis in order to relax the zero-order constraint in the VB approximation, further reducing the KLD of the optimal approximation.

Palm helped popularize the idea of a modern smartphone with its Palm Treo, but after failing to keep up with the iPhone and Android the company has been reborn. Its new product: A 3.3-inch companion device to work alongside those phones, not replace it. And yet Palm's past found me wanting to embrace the newly reborn Palm, an adorable and, yes, palm-sized $349 smartphone stand-in on sale Friday as a Verizon Wireless exclusive. I have fond memories for the original PalmPilot, the first truly successful personal digital assistant. I was even a fan of the Palm Pre, a phone with software that was, in hindsight, ahead of its time but couldn't escape the iPhone's shadow.

Pascal Weinberger in conversation with his team at Telefonica's "moonshots" division Alpha, where he heads AI research and development. Flying cars, augmented reality glasses and contact lenses that can detect diabetes: They're all innovations born out of Google X, the skunkworks division of Alphabet. Three years ago Spanish telco giant Telefónica established Alpha, a lab in Barcelona staffed by around 100 people, working in stealth on innovative technology that holds the promise of a potential new revenue streams. The person running all things AI at the lab is Pascal Weinberger, 21. Weinberger is originally from Germany and like many other computer programmers is self-taught.

If a recent demonstration in Canada is any indication, the digital vineyard of the future might very well receive scientific data in real-time gathered by drones and transferred via a cell network. Global UAV Technologies Ltd., Jöst Vineyards, VineView (Scientific Aerial Imaging Inc.) and a major Canadian telecommunications company recently completed a 4G proof-of-concept mission in in Malagash, Nova Scotia, to demonstrate a real-word application of drone technology for a project called the "Digital Vineyard of the Future." "Fine wine making is in the growing of grapes with specific qualities, where many variables have to be taken into consideration," said Jonathan Rodwell, director of viticulture and winemaking for Jöst Vineyards. "We see these emerging technologies offering excellent opportunities for integrated measurement and management of our vineyards and focus on precision viticulture." Global UAV provided a 4G-enabled, Procyon 800E helicopter drone platform with a specialized multi-spectral imaging payload.

Telefonica has become one of the first major technology companies to reveal exactly what guidelines it will place on the use of AI technology. The Spanish telecoms giant and parent company of O2 has released its Principles of Artificial Intelligence, laying out how it plans to utilise the technology to ensure a positive impact on society. Promising "integrity and transparency" in its design and development of AI technology, Telefonica's rules include the company's ongoing stance on areas such as equality, transparency, clarity, privacy, and security. "We're concerned about the possible use of artificial intelligence for the creation or dissemination of fake news, addiction to technology, and the possible reinforcement of social bias in the algorithms in general", said José María Álvarez-Pallete, Chairman & CEO of Telefónica. "These phenomena undermine the trust of our customers, our most valuable asset, and hinder the development of a fairer society. Consequently, we will do everything in our power to collaborate with other entities in order to eradicate them", he adds.

Huawei has made really good phones for years, but the Mate 20 Pro is the Chinese firm's first truly cutting-edge device with a triple camera, 3D face unlock and an in-screen fingerprint sensor. The Mate series of phones has always delivered one thing above all else – battery life. This year Huawei has gone out of its way to deliver even more. The Mate 20 Pro is the best feeling, most premium device the Chinese firm has made. It's incredibly solid, smooth and well built, but at 189g is still surprisingly light and manageable compared to the 208g iPhone XS Max and 201g Samsung Galaxy Note 9.

Graph neural networks (GNNs) have been shown to replicate convolutional neural networks' (CNNs) superior performance in many problems involving graphs. By replacing regular convolutions with linear shift-invariant graph filters (LSI-GFs), GNNs take into account the (irregular) structure of the graph and provide meaningful representations of network data. However, LSI-GFs fail to encode local nonlinear graph signal behavior, and so do regular activation functions, which are nonlinear but pointwise. To address this issue, we propose median activation functions with support on graph neighborhoods instead of individual nodes. A GNN architecture with a trainable multirresolution version of this activation function is then tested on synthetic and real-word datasets, where we show that median activation functions can improve GNN capacity with marginal increase in complexity.