Vigilante citizens in a town in Arizona have slashed tyres, thrown rocks and even pointed guns at self-driving vehicles being tested in their neighbourhood, an investigation has revealed. Police in Chandler recorded 21 incidents over the past two years in which the autonomous vehicles and their test drivers were targeted by local residents. One incident on 24 October saw a man emerge from a park and slash the tyres of a Waymo vehicle stopped at an intersection. Earlier this year a Waymo test driver reported a man in shorts aiming a gun at his car when it passed the man's driveway. Police reports also show that rocks were thrown at Waymo's fleet on at least four separate occasions, according to The Arizona Republic, while other incidents include people yelling at the vehicles, chasing them and forcing them off the road.

When you're trying to come up with gift ideas for someone who likes to cook, you want to find something that's both personal and practical. But finding a gift for a home cook that strikes that balance can be hard, especially if you're the kind of person whose fridge is filled with takeout containers. That's why we've gathered 35 of the best gifts for every type of home cook on your list--from the newbie who just wants to make a good grilled cheese to the home cook who has it all--all of them are available on Amazon, most of them with two-day Prime shipping. ChefSteps Joule Sous Vide, 1100 Watts, All White ($179) They might not think they need a sous vide machine, but that's exactly what makes it a great gift for an experienced chef, who can use it to make always-tender steaks, never-overcooked fish, and even soft-scrambled eggs. Echo Show (Second Generation) ($230) The new generation of the Echo Show has louder speakers and a bigger screen than before, so they can follow along with recipe videos and tutorials from any one of Amazon's partners, or ask Alexa to set a timer.

China is just days away from kicking off a historic mission to explore the dark side of the moon. The Chang'e-4, which launches on December 8th, will be the first ever rover to land on the far side of the lunar surface. The spacecraft is slated to leave from the Xichang Satellite Launch Center in Sichuan province, before entering the moon's orbit and hopefully touching down in the massive, 186-kilometer-wide Von Karman crater on the moon's south pole. The Chang'e-4, which launches on December 8th, will be the first ever rover land on the far side of the lunar surface. It will visit an unexplored region of the lunar surfacecalled the South Pole-Aitken Basin, located in the southern hemisphere of the dark side of the moon, which includes the Von Karman crater, among many others.

Deep generative models such as generative adversarial networks, variational autoencoders, and autoregressive models are rapidly growing in popularity for the discovery of new molecules and materials. In this work, we introduce MOlecular SEtS (MOSES), a benchmarking platform to support research on machine learning for drug discovery. MOSES implements several popular molecular generation models and includes a set of metrics that evaluate the diversity and quality of generated molecules. MOSES is meant to standardize the research on the molecular generation and facilitate the sharing and comparison of new models. Additionally, we provide a large-scale comparison of existing state of the art models and elaborate on current challenges for generative models that might prove fertile ground for new research. Our platform and source code are freely available at https://github.com/molecularsets/

Nov 28, 2018 Ginny Engholm Danny Iny, on How to Future-Proof Your Job Digital disruption is transforming the world of work. The skills that workers will need in the long term are changing at a rapid pace, and this is causing employees -- and employers -- a lot of anxiety. Workers fear being replaced by tech for good reason -- a 2017 report from PricewaterhouseCoopers estimated that 38 percent of U.S. jobs are at risk of being replaced by automation by the early 2030s. Some industries, such as transportation, wholesale, manufacturing and retail, are particularly vulnerable to automation. But no industry is safe.

Replanting trees after a wildfire or logging operation is an extremely labor intensive and expensive task. Carrying a bag of seedlings and using a dibble bar or shovel across steep debris-covered terrain can wear out a human. A new company, DroneSeed, has a solution. They are designing a system around a swarm of drones that can plant tree seeds in places where they have a decent chance of survival. First they survey the area with a drone using lidar and a multispectral camera to map the terrain and the vegetation.

Manifold embedding algorithms map high dimensional data, down to coordinates in a much lower dimensional space. One of the aims of the dimension reduction is to find the {\em intrinsic coordinates} that describe the data manifold. However, the coordinates returned by the embedding algorithm are abstract coordinates. Finding their physical, domain related meaning is not formalized and left to the domain experts. This paper studies the problem of recovering the domain-specific meaning of the new low dimensional representation in a semi-automatic, principled fashion. We propose a method to explain embedding coordinates on a manifold as {\em non-linear} compositions of functions from a user-defined dictionary. We show that this problem can be set up as a sparse {\em linear Group Lasso} recovery problem, find sufficient recovery conditions, and demonstrate its effectiveness on data.

We present a generative approach to classify scarcely observed longitudinal patient trajectories. The available time series are represented as tensors and factorized using generative deep recurrent neural networks. The learned factors represent the patient data in a compact way and can then be used in a downstream classification task. For more robustness and accuracy in the predictions, we used an ensemble of those deep generative models to mimic Bayesian posterior sampling. We illustrate the performance of our architecture on an intensive-care case study of in-hospital mortality prediction with 96 longitudinal measurement types measured across the first 48-hour from admission. Our combination of generative and ensemble strategies achieves an AUC of over 0.85, and outperforms the SAPS-II mortality score and GRU baselines.

Range Rover has unveiled its latest'green' compact luxury 4X4 with'X-Ray vision' that enable drivers to'see' through the bonnet to view what lies on beneath on the road below and'eyes in the back of its head' to show what's coming up behind. The new second-generation British-built Evoque off-roader is built with electrification and hybrid power in mind, and has a non-leather cloth option designed for vegetarians and vegans and made from recycled plastic bottles. Significantly, it features unique'ground view' technology that makes its bonnet'invisible' to the driver so he or she has a view of any rough or extreme terrain - or high city centre kerbs - that are lurking underneath or just ahead, but normally out of sight. The firm said its ClearSight Ground View technology marks'a world first' for the compact luxury 4X4 which is powered by a range of mild-hybrid petrol and diesel engines, with a plug in hybrid to follow next year and the potential for a fully electric version. Range Rover has unveiled its latest'green' compact luxury 4X4 with'X-Ray vision' that enable drivers to'see' through the bonnet to view what lies on beneath on the road below Significantly, it features unique'ground view' technology that makes its bonnet'invisible' to the driver so he or she has a view of any rough or extreme terrain - or high city centre kerbs - that are lurking underneath or just ahead, but normally out of sight The new all-seeing Evoque was unveiled at a star-studded international launch at the Old Truman Brewery in East London's trendy Brick Lane where fashion and Vogue cover model Adwoa Aboah was to appear wearing a dress made out of the very same Kvadrat material made from recycled plastic bottles.

Keyword spotting--or wakeword detection--is an essential feature for hands-free operation of modern voice-controlled devices. With such devices becoming ubiquitous, users might want to choose a personalized custom wakeword. In this work, we present DONUT, a CTC-based algorithm for online query-by-example keyword spotting that enables custom wakeword detection. The algorithm works by recording a small number of training examples from the user, generating a set of label sequence hypotheses from these training examples, and detecting the wakeword by aggregating the scores of all the hypotheses given a new audio recording. Our method combines the generalization and interpretability of CTC-based keyword spotting with the user-adaptation and convenience of a conventional query-by-example system. DONUT has low computational requirements and is well-suited for both learning and inference on embedded systems without requiring private user data to be uploaded to the cloud.