11 January 2018

Intel details performance hit for Meltdown fix on affected processors


 Now that the Meltdown and Spectre cats are out of the bag and patches are being issued left and right, the main question people have is: will this affect my everyday work? Intel’s latest performance metrics suggest that yes it might, but not too much — more so if, like me, you have an older processor. Read More

Read Full Article

9 Great Third-Party Amazon Alexa and Google Assistant Smart Speakers


third-party-alexa-google-speakers

The smart speaker concept has come a long way since 2014 when the first Amazon Echo arrived. Today, the online retailer offers no less than six Amazon Alexa-based speakers, including a second-generation Echo, premium Echo Plus, and fashionable Echo Show. Thanks to Amazon’s success, Google now also offers smart speakers powered by Google Assistant. These speakers include the 2016 Google Home, 2017 Google Home Mini, and 2017 Google Home Max. But these aren’t the only Amazon Alexa- and Google Assistant-based speakers on the market. Thanks to public SDKs, other hardware companies are also getting in on the game. As such,...

Read the full article: 9 Great Third-Party Amazon Alexa and Google Assistant Smart Speakers


Read Full Article

The Google Brain Team — Looking Back on 2017 (Part 1 of 2)




The Google Brain team works to advance the state of the art in artificial intelligence by research and systems engineering, as one part of the overall Google AI effort. Last year we shared a summary of our work in 2016. Since then, we’ve continued to make progress on our long-term research agenda of making machines intelligent, and have collaborated with a number of teams across Google and Alphabet to use the results of our research to improve people’s lives. This first of two posts will highlight some of our work in 2017, including some of our basic research work, as well as updates on open source software, datasets, and new hardware for machine learning. In the second post we’ll dive into the research we do in specific domains where machine learning can have a large impact, such as healthcare, robotics, and some areas of basic science, as well as cover our work on creativity, fairness and inclusion and tell you a bit more about who we are.

Core Research
A significant focus of our team is pursuing research that advances our understanding and improves our ability to solve new problems in the field of machine learning. Below are several themes from our research last year.

AutoML
The goal of automating machine learning is to develop techniques for computers to solve new machine learning problems automatically, without the need for human machine learning experts to intervene on every new problem. If we’re ever going to have truly intelligent systems, this is a fundamental capability that we will need. We developed new approaches for designing neural network architectures using both reinforcement learning and evolutionary algorithms, scaled this work to state-of-the-art results on ImageNet classification and detection, and also showed how to learn new optimization algorithms and effective activation functions automatically. We are actively working with our Cloud AI team to bring this technology into the hands of Google customers, as well as continuing to push the research in many directions.
Convolutional architecture discovered by Neural Architecture Search
Object detection with a network discovered by AutoML
Speech Understanding and Generation
Another theme is on developing new techniques that improve the ability of our computing systems to understand and generate human speech, including our collaboration with the speech team at Google to develop a number of improvements for an end-to-end approach to speech recognition, which reduces the relative word error rate over Google’s production speech recognition system by 16%. One nice aspect of this work is that it required many separate threads of research to come together (which you can find on Arxiv: 1, 2, 3, 4, 5, 6, 7, 8, 9).
Components of the Listen-Attend-Spell end-to-end model for speech recognition
We also collaborated with our research colleagues on Google’s Machine Perception team to develop a new approach for performing text-to-speech generation (Tacotron 2) that dramatically improves the quality of the generated speech. This model achieves a mean opinion score (MOS) of 4.53 compared to a MOS of 4.58 for professionally recorded speech like you might find in an audiobook, and 4.34 for the previous best computer-generated speech system. You can listen for yourself.
Tacotron 2’s model architecture
New Machine Learning Algorithms and Approaches
We continued to develop novel machine learning algorithms and approaches, including work on capsules (which explicitly look for agreement in activated features as a way of evaluating many different noisy hypotheses when performing visual tasks), sparsely-gated mixtures of experts (which enable very large models that are still computational efficient), hypernetworks (which use the weights of one model to generate weights for another model), new kinds of multi-modal models (which perform multi-task learning across audio, visual, and textual inputs in the same model), attention-based mechanisms (as an alternative to convolutional and recurrent models), symbolic and non-symbolic learned optimization methods, a technique to back-propagate through discrete variables, and a few new reinforcement learning algorithmic improvements.

Machine Learning for Computer Systems
The use of machine learning to replace traditional heuristics in computer systems also greatly interests us. We have shown how to use reinforcement learning to make placement decisions for mapping computational graphs onto a set of computational devices that are better than human experts. With other colleagues in Google Research, we have shown in “The Case for Learned Index Structures” that neural networks can be both faster and much smaller than traditional data structures such as B-trees, hash tables, and Bloom filters. We believe that we are just scratching the surface in terms of the use of machine learning in core computer systems, as outlined in a NIPS workshop talk on Machine Learning for Systems and Systems for Machine Learning.
Learned Models as Index Structures
Privacy and Security
Machine learning and its interactions with security and privacy continue to be major research foci for us. We showed that machine learning techniques can be applied in a way that provides differential privacy guarantees, in a paper that received one of the best paper awards at ICLR 2017. We also continued our investigation into the properties of adversarial examples, including demonstrating adversarial examples in the physical world, and how to harness adversarial examples at scale during the training process to make models more robust to adversarial examples.

Understanding Machine Learning Systems
While we have seen impressive results with deep learning, it is important to understand why it works, and when it won’t. In another one of the best paper awards of ICLR 2017, we showed that current machine learning theoretical frameworks fail to explain the impressive results of deep learning approaches. We also showed that the “flatness” of minima found by optimization methods is not as closely linked to good generalization as initially thought. In order to better understand how training proceeds in deep architectures, we published a series of papers analyzing random matrices, as they are the starting point of most training approaches. Another important avenue to understand deep learning is to better measure their performance. We showed the importance of good experimental design and statistical rigor in a recent study comparing many GAN approaches that found many popular enhancements to generative models do not actually improve performance. We hope this study will give an example for other researchers to follow in making robust experimental studies.

We are developing methods that allow better interpretability of machine learning systems. And in March, in collaboration with OpenAI, DeepMind, YC Research and others, we announced the launch of Distill, a new online open science journal dedicated to supporting human understanding of machine learning. It has gained a reputation for clear exposition of machine learning concepts and for excellent interactive visualization tools in its articles. In its first year, Distill has published many illuminating articles aimed at understanding the inner working of various machine learning techniques, and we look forward to the many more sure to come in 2018.
Feature Visualization
How to Use t-SNE effectively
Open Datasets for Machine Learning Research
Open datasets like MNIST, CIFAR-10, ImageNet, SVHN, and WMT have pushed the field of machine learning forward tremendously. Our team and Google Research as a whole have been active in open-sourcing interesting new datasets for open machine learning research over the past year or so, by providing access to more large labeled datasets including:
Examples from the YouTube-Bounding Boxes dataset: Video segments sampled at 1 frame per second, with bounding boxes successfully identified around the items of interest.
TensorFlow and Open Source Software
A map showing the broad distribution of TensorFlow users (source)
Throughout our team’s history, we have built tools that help us to conduct machine learning research and deploy machine learning systems in Google’s many products. In November 2015, we open-sourced our second-generation machine learning framework, TensorFlow, with the hope of allowing the machine learning community as a whole to benefit from our investment in machine learning software tools. In February, we released TensorFlow 1.0, and in November, we released v1.4 with these significant additions: Eager execution for interactive imperative-style programming, XLA, an optimizing compiler for TensorFlow programs, and TensorFlow Lite, a lightweight solution for mobile and embedded devices. The pre-compiled TensorFlow binaries have now been downloaded more than 10 million times in over 180 countries, and the source code on GitHub now has more than 1,200 contributors.

In February, we hosted the first ever TensorFlow Developer Summit, with over 450 people attending live in Mountain View and more than 6,500 watching on live streams around the world, including at more than 85 local viewing events in 35 countries. All talks were recorded, with topics ranging from new features, techniques for using TensorFlow, or detailed looks under the hoods at low-level TensorFlow abstractions. We’ll be hosting another TensorFlow Developer Summit on March 30, 2018 in the Bay Area. Sign up now to save the date and stay updated on the latest news.
This rock-paper-scissors science experiment is a novel use of TensorFlow. We’ve been excited by the wide variety of uses of TensorFlow we saw in 2017, including automating cucumber sorting, finding sea cows in aerial imagery, sorting diced potatoes to make safer baby food, identifying skin cancer, helping to interpret bird call recordings in a New Zealand bird sanctuary, and identifying diseased plants in the most popular root crop on Earth in Tanzania!
In November, TensorFlow celebrated its second anniversary as an open-source project. It has been incredibly rewarding to see a vibrant community of TensorFlow developers and users emerge. TensorFlow is the #1 machine learning platform on GitHub and one of the top five repositories on GitHub overall, used by many companies and organizations, big and small, with more than 24,500 distinct repositories on GitHub related to TensorFlow. Many research papers are now published with open-source TensorFlow implementations to accompany the research results, enabling the community to more easily understand the exact methods used and to reproduce or extend the work.

TensorFlow has also benefited from other Google Research teams open-sourcing related work, including TF-GAN, a lightweight library for generative adversarial models in TensorFlow, TensorFlow Lattice, a set of estimators for working with lattice models, as well as the TensorFlow Object Detection API. The TensorFlow model repository continues to grow with an ever-widening set of models.

In addition to TensorFlow, we released deeplearn.js, an open-source hardware-accelerated implementation of deep learning APIs right in the browser (with no need to download or install anything). The deeplearn.js homepage has a number of great examples, including Teachable Machine, a computer vision model you train using your webcam, and Performance RNN, a real-time neural-network based piano composition and performance demonstration. We’ll be working in 2018 to make it possible to deploy TensorFlow models directly into the deeplearn.js environment.

TPUs
Cloud TPUs deliver up to 180 teraflops of machine learning acceleration
About five years ago, we recognized that deep learning would dramatically change the kinds of hardware we would need. Deep learning computations are very computationally intensive, but they have two special properties: they are largely composed of dense linear algebra operations (matrix multiples, vector operations, etc.), and they are very tolerant of reduced precision. We realized that we could take advantage of these two properties to build specialized hardware that can run neural network computations very efficiently. We provided design input to Google’s Platforms team and they designed and produced our first generation Tensor Processing Unit (TPU): a single-chip ASIC designed to accelerate inference for deep learning models (inference is the use of an already-trained neural network, and is distinct from training). This first-generation TPU has been deployed in our data centers for three years, and it has been used to power deep learning models on every Google Search query, for Google Translate, for understanding images in Google Photos, for the AlphaGo matches against Lee Sedol and Ke Jie, and for many other research and product uses. In June, we published a paper at ISCA 2017, showing that this first-generation TPU was 15X - 30X faster than its contemporary GPU or CPU counterparts, with performance/Watt about 30X - 80X better.
Cloud TPU Pods deliver up to 11.5 petaflops of machine learning acceleration
Experiments with ResNet-50 training on ImageNet show near-perfect speed-up as the number of TPU devices used increases.
Inference is important, but accelerating the training process is an even more important problem - and also much harder. The faster researchers can try a new idea, the more breakthroughs we can make. Our second-generation TPU, announced at Google I/O in May, is a whole system (custom ASIC chips, board and interconnect) that is designed to accelerate both training and inference, and we showed a single device configuration as well as a multi-rack deep learning supercomputer configuration called a TPU Pod. We announced that these second generation devices will be offered on the Google Cloud Platform as Cloud TPUs. We also unveiled the TensorFlow Research Cloud (TFRC), a program to provide top ML researchers who are committed to sharing their work with the world to access a cluster of 1,000 Cloud TPUs for free. In December, we presented work showing that we can train a ResNet-50 ImageNet model to a high level of accuracy in 22 minutes on a TPU Pod as compared to days or longer on a typical workstation. We think lowering research turnaround times in this fashion will dramatically increase the productivity of machine learning teams here at Google and at all of the organizations that use Cloud TPUs. If you’re interested in Cloud TPUs, TPU Pods, or the TensorFlow Research Cloud, you can sign up to learn more at g.co/tpusignup. We’re excited to enable many more engineers and researchers to use TPUs in 2018!

Thanks for reading!

(In part 2 we’ll discuss our research in the application of machine learning to domains like healthcare, robotics, different fields of science, and creativity, as well as cover our work on fairness and inclusion.)

Boeing’s prototype drone can carry 500 lbs of cargo


 Boeing just revealed a prototype drone capable of carrying much more than a camera. The company tasked engineers with designing and building a cargo drone and the prototype they came up with is able to haul 500 lbs of goods. The vehicle is huge and much larger than anything DJI sells. It weighs 747 pounds and is 15 feet long, 18 feet wide and 4 feet tall. Four arms hold two props each. It… Read More

Read Full Article

Avoid IRS Scams: 7 Warning Signs to Watch Out For


irs-scams

Now that we’ve entered the new year, U.S. residents are starting to think about their taxes. Scammers know this, and they’re out to cheat you out of money. IRS scams are commonplace, because scammers know how nervous people are to deal with the IRS. But if you know how to recognize tax scams, you can stay safe throughout tax season. There are a few things you need to watch out for. Here’s how to avoid being fooled by a scammer impersonating the IRS. 1. The IRS Won’t Call First How the IRS contacts you depends on a few different factors,...

Read the full article: Avoid IRS Scams: 7 Warning Signs to Watch Out For


Read Full Article

LiveMap shows off latest prototype of augmented reality motorcycle helmet


 LiveMap launched in TechCrunch’s Hardware Battlefield CES 2014 and is getting closer to producing its augmented reality helmet. Since launching, the product has evolved to a working prototype. I tried it on at CES and it seemed to work fine. The current version of helmet features a small transparent screen mounted on the visor. A small projector is mounted in the chin of the helmet.… Read More

Read Full Article

7 Ways You Can Help Save YouTube From Itself


save-youtube

YouTube is the internet’s home for videos, but it had rather a rough 2017. After a scare concerning their ads being seen before inappropriate content, advertisers removed their commercials from videos en masse. This was known as the “adpocalypse” and meant that creators everywhere lost money. Loads of videos are now flagged as ineligible for monetization for seemingly no reason. The YouTube Kids app has been inundated with creepy videos masquerading as children’s cartoons. What start out as videos featuring characters from Peppa Pig, Frozen, and other popular kid shows turn into violent, inappropriate, and terrifying scenes. And many YouTube...

Read the full article: 7 Ways You Can Help Save YouTube From Itself


Read Full Article

Google acquired Redux, a U.K. startup focused on audio and haptics


 Alphabet, the umbrella corporation of Google, Inc. etc., has quietly acquired a UK-based startup called Redux, reports Bloomberg. Redux was founded in 2013 out of Cambridge, and built technology that uses vibrations to turn surfaces of phones or tablets into speakers or provide haptic feedback. The acquisition is reflected on Crunchbase, and in confirmed transfer of shares within U.K.… Read More
Read Full Article

These 30-Day Learning Challenges Can Completely Change Your Year


30day-learning-challenges

Starts with a bang. Ends with a whimper. That’s the story of most new year resolutions. But, challenge yourself to learn a new skill this year. One of the solutions to your new year resolution woes according to the American Psychological Association is to start small. They also say: Change one behavior at a time Learn a new skill one at a time. Maybe, don’t set the horizon too far away. Combine the three tenets of small starts, short time frames, and one new skill or habit at a time into brief challenges of 30 days or less. A month can be just the...

Read the full article: These 30-Day Learning Challenges Can Completely Change Your Year


Read Full Article

Voice Dictation – Type with your Voice


Introducing the all-new Voice Dictation v2.0, a speech recognition app that lets you type with your voice. There’s no software to install, there’s no training required and all you need is Google Chrome on your Windows PC, Mac OS or Linux.

Dictation can recognize spoken words in English, Hindi, Español, Italiano, Deutsch, Français, and all the other popular languages. Another unique feature of Dictation is support for voice commands that let you do more with your voice. For instance, you can say a command like new line or nueva línea for inserting lines. You can add punctuations, special symbols and even smileys using simple commands in most languages.

This YouTube video will walk you through the Dictation app.

Open Dictation

How to use Dictation for Speech Recognition

Dictation stores everything in your browser locally and not a byte of your data is uploaded anywhere.  Speech recognition will be more accurate in a quiet environment and the built-in microphone of your computer should be good enough.

Open dictation.io in Google Chrome and choose your native language from the drop-down. Then click the microphone icon (or tap the Start button) to activate web speech and start speaking.

The first time you use Dictation, you would need to allow permissions to the website to access your microphone.  Also, make sure that no other browser tab is using Speech Recognition at the same time.

allow-microphone-access.png

As you speak, your voice is continuously converted into text in real time. There are no limits o and when you are done with dictation, you can click the “Stop” button or say “Go to Sleep” to turn off speech recognition.

Speech to Text with Dictation

Dictation supports speech to text as well so you can listen to the transcribed audio in any of the available voices. You can alter the speed and pitch of the playback with simple controls.

Publish to the Web

Dictation includes a WYSIWYG editor to help you format the transcribed text. You can copy the rich-text to the clipboard with a click and paste into other apps like Gmail or Microsoft Word with the formatting.

You also have the “Publish” button to anonymously publish your note to the Internet and this page can only be accessed by users who know the link.

dictation.voice.png

Convert Recorded Audio to Text

If you have an MP3 file that you would like to transcribe to text, Dictation can help you do that well. Choose the language on dictation.io, click the start button, then play the audio file on your mobile and watch as the words are converted into text in real time. See demo.

Dictation – The Technical Stuff

Dictation uses the HTML5 Web Speech API that is currently implemented only in Google Chrome on the desktop. For Text to Speech, it uses the SpeechSynthesis interface of the Web Speech API that is available in Chrome, Microsoft Edget and Firefox.

The rich-text editor is built with Quill while the voice command interface is made possible with Annyang. Dictation uses the Github API to publish dictated notes online as anonymous gists.


How Not to Manage Your To-Do List: 8 Mistakes You Must Avoid


manage-todo-lists

You may have read countless blog posts, magazine articles, and books about to-do lists. They certainly exist, but many of the articles you read often present them as the magic tools necessary for enjoying consistent productivity. They could be that magic tool but, the thing is, merely making a to-do list is not enough. Many people unknowingly make mistakes that limit the overall effectiveness of their to-do lists. The blunders may even hinder their potential. Below, we’ll look at some of the most common mistakes associated with to-do lists. You’ll also get solutions to these mistakes, so you can remedy...

Read the full article: How Not to Manage Your To-Do List: 8 Mistakes You Must Avoid


Read Full Article

Why You Should Try Switching Linux Operating Systems


try-switching-linux-os

If you’ve ever felt your current Linux operating system lacking in some ways, there are a number of things you can do to help fix this. It’s easy to adjust your desktop’s look and feel, for example, sometimes to amazing degrees. Switching regularly between different Linux operating systems (better known as distro hopping) is one potential solution, which carries a number of benefits. What Is It? Some people who use Linux enjoy moving between distributions, for varied, and personal reasons. Even though at their core, the base is the same, there’s still a lot that can change between them. And...

Read the full article: Why You Should Try Switching Linux Operating Systems


Read Full Article

The Google Docs Guide You Need for All Your Business Documents


google-docs-guide-business

You have many tools for creating and sharing documents. Report builders, document editors, and email applications all play a part in your daily duties. But have you ever considered Google Docs in earnest? Google’s own office productivity suite has features that can help you create, edit, and share your documents quickly and easily. Wherever you are and whatever your work, Google Docs can save you time — and you don’t have to spend a dime. Let’s see the helpful Google Docs features that can work for you and your team. Use Google Docs for Business Collaboration When you work in...

Read the full article: The Google Docs Guide You Need for All Your Business Documents


Read Full Article

ECJ to rule on whether Facebook needs to hunt for hate speech


 Austria’s Supreme Court is referring a legal challenge over the extent of Facebook’s responsibility to remove hate speech postings to Europe’s top court for an opinion. The case has clear implications for freedom of speech online. Read More
Read Full Article

You Can Still Upgrade to Windows 10 for Free!


Windows 10 is not perfect, but it’s a significant improvement on everything that’s gone before, mainly thanks to the three massive upgrades it’s received: the Anniversary Update, Creators Update, and the Fall Creators Update. Unfortunately, if you’re still using Windows 7 or 8, you’ve missed the initial free update period and you’re now cursed to be stuck in the past forever… Or are you? Actually, no! It turns out you can still grab a free upgrade to Windows 10 using your old product key. Yes, the Trick Still Works! Technically, the period for free upgrades closed more than 12 months ago. However, throughout...

Read the full article: You Can Still Upgrade to Windows 10 for Free!


Read Full Article

How to Disable OneNote From Pasting Source Links


onenote-wiki

If you’re a regular OneNote user, you’ve probably come across this annoying feature: when pasting text from a website, OneNote automatically includes the link to that site with your copied text. It’s worth mentioning that the feature does have its uses. For example, if you’re collecting research and want to be sure to cite all of your sources, this feature can be invaluable — a great way to be sure that you’re giving credit where credit is due. If however, hanging on to those links is unnecessary, it just adds another step for you to do when formatting your OneNote documents. If you...

Read the full article: How to Disable OneNote From Pasting Source Links


Read Full Article

How to Gameshare on Xbox One


Gamesharing with a trusted friend is a great way to save money on gaming. You and a buddy can take turns buying games and both have access to them by switching your home Xbox. If you want to gameshare on your Xbox One, we’ll show you the process and note a few important points. How to Gameshare on Xbox One Find a trusted friend whom you want to share games with. Either obtain their Xbox Live login information, or invite them over so they can sign into your console. Press the Xbox button on your controller to open the Guide....

Read the full article: How to Gameshare on Xbox One


Read Full Article

How to Learn a New Language by Reading the Daily News


The best way to learn a new language is to immerse yourself in it. But when you don’t have a speaking partner or aren’t in the native country, then you have to come up with another idea. How about practicing with real-world sentences in the vocabulary of everyday conversation? There are several ways to do that, but this unique language-learning web app teaches you with the help of the news — one sentence at a time, in short bursts. Learn a Language by Reading the Daily News WordBrewery scans newspapers and gives you the right sentences to learn from. In their own words, the news...

Read the full article: How to Learn a New Language by Reading the Daily News


Read Full Article

YouTube drops Logan Paul from Google Preferred and puts his Originals on hold


 YouTube has taken further action against social media star Logan Paul, dropping the vlogger from its Google Preferred program, which is meant to be a mark of trust to signal to advertisers they can rely on these media creators to generate higher quality content. After Paul posted a video of a dead body he filmed hanging from a tree in Japan’s colloquially titled ‘suicide… Read More
Read Full Article

Dell’s new app brings mobile notifications to the desktop


 I’ll be honest, Dell wasn’t at the top of our list of must visit companies at this year’s CES, but the PC stalwart is actually showcasing a couple of interesting new products at the show. On the software front, mobile connect is a free offering that will be bundled with all new Dell computers. The application delivers mobile notifications to the desktop via Bluetooth.… Read More

Read Full Article

YouTube’s in-app messaging and Community tab to make their way to YouTube TV, YouTube Music


 YouTube is aiming to bring its set of social features, including the in-app messaging system and “Community” tab for creators, to its wider suite of apps. Specifically, the company is interested in porting those features to its YouTube TV app aimed at cord cutters, as well as its Music app. The company won’t confirm a timeline in terms of when these features would launch,… Read More
Read Full Article