Monthly:December 2016

Bringing the best of mobile technology to Imperial College Healthcare NHS Trust

Were really excited to announce that weve agreed a five year partnership with Imperial College Healthcare NHS Trust, helping them make the most of the opportunity for mobile clinical applications to improve care. This is now our second NHS partnership for clinical apps, following a similar partnership we announced last month with the Royal Free London NHS Foundation Trust.Over the last two years, the Trust has moved from paper to electronic patient records, and mobile technology is the natural next stage of this work. By giving clinicians access to cutting-edge healthcare apps that link to electronic patient records, theyll be able to access information on the move, react quickly in response to changing patient needs, and ultimately provide even better care.Well be working with the Trust to deploy our clinical app, Streams, which supports clinicians in caring for patients at risk of deterioration, particularly with conditions where early intervention can make all the difference. Like breaking news alerts on a mobile phone, the technology will notify nurses and doctors immediately when test results show a patient is at risk of becoming seriously ill. It will also enable clinicians at the Trust to securely assign and communicate about clinical tasks, and give them the information they need to make diagnoses and decisions.Read More

(Google) Home for the Holidays

It’s that time of year. Holiday music, sweet treats, classic movies — all the reasons to cozy up and spend time at home with the ones you love. And with Google Home, there are countless ways to spread the holiday cheer.  

Play Netflix on your TV with the magic words “Ok Google”

Serve up the holiday classics that make your season bright, hands-free, right to your TV — when Google Home is paired with a Chromecast device. Now, in addition to YouTube videos, you can stream any movie or TV series from Netflix to the TV, using just your voice*. Start with “Ok Google, play Stranger Things from Netflix on my TV” or “Ok Google, play Love Actually on Netflix on my TV”. You can even control your media with commands like “Pause this episode”, without lifting a finger. Here’s how to link your Netflix account to the Google Assistant from the Google Home app (make sure you’re running the latest version). Netflix linking in the Google Home app is available now for iOS and rolling out to Android this week.

Enjoy your Google Photos on the big screen, hands-free.

The holidays is the perfect time to gather around and relive memories from the year. Google Photos now works with Google Home, so you can share your photos on your Chromecast-connected TV, using just your voice. Simply say, “Ok Google, show my photos of Lake Tahoe on my TV” to see your beautiful pictures bigger. Even if you can’t remember the name of the album, you can search for pictures. Try “Ok Google, show my photos of selfies at Christmas on my TV” or “Ok Google, show my photos of Brazil sunsets on Chromecast”. You can even use your voice to search for your places, things and people.

Learn how to link your Photos account to the Google Assistant here.


Enjoy holiday music from your favorite apps

Fill your home with festive tunes from your favorite music apps like Google Play Music, Spotify, Pandora, or YouTube Music with just a simple voice command. Say, “Ok Google, play Hanukkah music” or “Ok Google, play Christmas music” and you’ll get into the spirit in no time.

Keep track of Santa

Whether he’s preparing his sleigh or already in flight, your Assistant keeps you informed of Santa’s latest whereabouts (tracked by Google Maps). Just say “Ok Google, where’s Santa right now?”

Get YouTube Red FREE for six months

Keep the holiday glow going well into the new year as you enjoy a premium music experience — more songs, more versions, no ads — with YouTube Red. Get six months free of YouTube Red when you buy a Google Home before December 31, 2016**.

Make your holidays truly magical with Google Home.

*Netflix streaming membership required

**Offer expires 12/31/16. Terms apply. New subscribers only.

Read More

Android Wear: The Magic Minute Project

What’s possible in a minute? You can fall in love. Launch a rocket ship. Set a rap world record.

Today, Android Wear is teaming up with with makers, doers, and dreamers to launch The Magic Minute Project. It’s a collection of one-minute films celebrating what time means to all of us—told one #MagicMinute at a time.

Andrew Huang: 300 Word Rap

A music and video producer, Andrew Huang has a penchant for working within absurd, self-imposed limitations. Using his Android Wear watch to time his performance, Andrew attempts to rap 300 words in one #MagicMinute.

Bruna Kajiya: Top of Your Sport

Bruna Kajiya is a Brazilian kiteboarder and World Champion and the first woman to land a double handle pass in kiteboarding. The final minute of a competition is always the most intense: The riders perform their “all or nothing” tricks and everything is on the line.

Kie Willis: Time Trial

Time trials have become popular in Parkour. Setting a route across an urban landscape, leaping between rooftops—it’s seen as the highest test of practice. Kie Willis, one of the most well known freerunners in the world, attempts a one-minute Parkour while keeping his heart rate low.

Zanita Whittington: Mix Up, Look Sharp

Putting together the perfect look takes time, but a versatile accessory can help. Zanita Whittington—model, photographer, and creative director—completes her look by personalizing her Android Wear watch face.

Mike Relm: OK Scratch

Mike Relm is a world renowned turntablist and director, known for sampling music in surprisingly innovative ways to create new sounds. He has sampled and scratched many sounds in his life, but this is the first time using his wife’s voice!

Just as traditional watches help tell the time, Android Wear watches help make the most of our time. Whatever makes a minute magical to you, why not share it with the world too? It can be anything you imagine: paint, play, code, cook, rap or render.

Submit your own #MagicMinute for a chance to be featured in The Magic Minute Project film. Follow @AndroidWear on Twitter or visit to explore the best #MagicMinute videos.

Read More

DeepMind Papers @ NIPS (Part 3)

Scaling Memory-Augmented Neural Networks with Sparse Reads and WritesAuthors:J Rae, JJ Hunt, T Harley, I Danihelka, A Senior, G Wayne, A Graves, T LillicrapWe can recall vast numbers of memories, making connections between superficially unrelated events. As you read a novel, youll likely remember quite precisely the last few things youve read, but also plot summaries, connections and character traits from far back in the novel.Many machine learning models of memory, such as Long Short Term Memory, struggle at these sort of tasks. The computational cost of these models scales quadratically with the number of memories they can store so they are quite limited in how many memories they can have. More recently, memory augmented neural networks such as the Differentiable Neural Computer or Memory Networks, have shown promising results by adding memory separate from the computation and solving tasks such as reading short stories and answering questions [e.g. Babi].However, while these new architectures show promising results on small tasks, they use “soft-attention for accessing their memories, meaning that at every timestep they touch every word in memory. So while they can scale to short stories, theyre a long way from reading novels.In this work, we develop a set of techniques to use sparse approximations of such models to dramatically improve their scalability.Read More

Hassle-free Wi-Fi with Network Assist

Wi-Fi has become an essential part of daily home life, and yet, for something that’s so critical, Wi-Fi is often forgotten. That is, until it stops working. We’ve all experienced a spinning buffering wheel when trying to stream a movie, or a room that just never quite seems to have a strong signal. And when something goes wrong, most of us are at a loss for how to fix it, crossing our fingers in hopes that turning the router off, then on again is the solution.

You shouldn’t have to be a network engineer to have good, reliable Wi-Fi at home, which is why we built Google Wifi, available for sale today at retailers like the Google Store, Amazon, Best Buy and Walmart.

One of the ways Google Wifi is different from your traditional router is its Network Assist technology. Network Assist is based on Google intelligence that manages all the complexities of Wi-Fi behind the scenes to make sure you always have the best connection possible. Network Assist helps keep your Google Wifi system running smoothly in three key ways:

Keeps your Wi-Fi network at peak performance

Most routers scan the environment for a clear channel when they are powered on, and never again (until you restart it, which is why it’s so common that “unplugging” seems to be the fix). Over time, Wi-Fi channels can get congested, slowing channel speeds to a crawl. But now, Google Wifi solves the congestion issue by performing a scan of the wireless spectrum every five minutes. Network Assist uses these results to build a predictive model to determine the best channel for your Wi-Fi to make sure you’re always on the clearest channel, which means less congestion and faster speeds.

To ensure you always have strong coverage throughout your home, Network Assist continuously monitors the connection between all your Google Wifi points to make sure they have good signal with each other. It also conducts regular speed tests in the background, so you can review your history to see what speeds you’re getting from your internet service provider.

Finds the best connection for your devices

Through Network Assist, you’ll always get the best Wi-Fi connectivity possible, but we take it a step further to make sure that each individual device always has the best connection.

Your devices (like laptops, smartphones, connected light bulbs) connect to Wi-Fi using one of two bands: 2.4GHz or 5GHz. The 2.4GHz band has slower speeds but a longer range, while 5GHz is faster but with shorter range. Network Assist helps your devices connect to the best band, based on the device and its location relative to Wifi points to make sure you always have a strong, reliable Wi-Fi signal.

Devices are getting smaller and more mobile — they move around with us. So Network Assist makes sure your devices seamlessly transition between Wifi points in your home as you move from room to room, so that you don’t get any drops in signal. Network Assist transitions you, while ensuring you continue to have a solid connection, so you can walk throughout the house on a video call.


Keeps you in the know

Traditional routers communicate through blinking lights, but Network Assist tells you about any issues in human language, right in the Google Wifi app. For example, it can let you know if you would get faster speeds by moving one of your Google Wifi points to another location in your home, or sends you a notification if you’ve lost connection to your Internet Service Provider (ISP). Getting these insights and real-time optimizations is as simple as opening up your Google Wifi app.

network assist

Wi-Fi should be easy to set up, provide great coverage for your entire house, and do all the heavy lifting for you. We built Google Wifi to answer those needs and more, and hope it improves your Wi-Fi experience at home.

Read More

DeepMind Papers @ NIPS (Part 2)

The second blog post in this series, sharing brief descriptions of the papers we are presenting at NIPS 2016 Conference in Barcelona.Sequential Neural Models with Stochastic LayersAuthors:Marco Fraccaro, Sren Kaae Snderby, Ulrich Paquet, Ole WintherMuch of our reasoning about the world is sequential, from listening to sounds and voices and music, to imagining our steps to reach a destination, to tracking a tennis ball through time. All these sequences have some amount of latent random structure in them. Two powerful and complementary models, recurrent neural networks (RNNs) and stochastic state space models (SSMs), are widely used to model sequential data like these. RNNs are excellent at capturing longer-term dependencies in data, while SSMs model uncertainty in the sequence’s underlying latent random structure, and are great for tracking and control.Is it possible to get the best of both worlds? In this paper we show how you can, by carefully layering deterministic (RNN) and stochastic (SSM) layers. We show how you can efficiently reason about a sequences present latent structure, given its past (filtering) and also its past and future (smoothing).For further details and related work, please see the paper it out at NIPS:Tue Dec 6th 05:20 – 05:40 PM @ Area 1+2 (Oral) in Deep LearningTue Dec 6th 06:00 – 09:30 PM @ Area 5+6+7+8 #179Read More

Open-sourcing DeepMind Lab

DeepMind’s scientific mission is to push the boundaries of AI, developing systems that can learn to solve any complex problem without needing to be taught how. To achieve this, we work from the premise that AI needs to be general. Agents should operate across a wide range of tasks and be able to automatically adapt to changing circumstances. That is, they should not be pre-programmed, but rather, able to learn automatically from their raw inputs and reward signals from the environment. There are two parts to this research program: (1) designing ever-more intelligent agents capable of more-and-more sophisticated cognitive skills, and (2) building increasingly complex environments where agents can be trained and evaluated.Read More

DeepMind Papers @ NIPS (Part 1)

Over the next three blogposts, we’re going to share with you brief descriptions of the papers we are presenting at the NIPS 2016 Conference in Barcelona.Read More