Two and a half months in advance to apply for an invitation code, one day in advance to collect badges and booking seats, 2.5 hours before the queue waiting, finally in the opening an hour before when we entered the Google I/O 2017 Keynote main venue.
Continuous screaming and cheering, self timer and photo visible everywhere, even in the first moment, we can feel the most attractive in the World Congress of science and technology field is what is full of excitement. The initial DJ, disc, and balloon games may not be to your taste, but that's OK. The two hours of high quality content output proves that this may still be the most valuable technology conference this year.
Here we will help you to sort out what Keynote links this morning, at about 4 minutes of video or 10 minutes of reading time, you will have a complete understanding of the two hours.
Google Lens: new AI vision search applications
First of all, Google CEO Sundar Pichai has released artificial intelligence applications Google Lens based on image recognition and OCR technology on I/O.
According to Pichai at the conference, the new technology will be used first on Google, Assistant and Google Photos, and will have a wider range of applications in the future. The point of Google Lens is that it allows smart phones to be more than justseeingThe world, it canReadEverything in the camera, and help you take some action.
In the live demonstration, Google demonstrated some of the applications of the new technology:
The simplest thing you can do is to identify objects, including flowers and other plant species.
Google Lens and some of the more intelligent functions, such as mobile phone Wi-Fi, a camera can be aligned on the router, Wi-Fi Google Lens will be able to automatically extract account and password, and then provide a key connection options.
In addition, it can intelligently identify the Street merchants and automatically retrieve the information.
In addition, in Google Photos applications, Google Lens can also help identify buildings and other content in the image, and can do simple image text recognition.
In fact, Google Translate has a similar function in the mobile phone application, application of Google in Translate, open the camera alignment a English can get similar font Chinese translation in the camera interface.
However, although the official clear Google Lens will be first equipped with voice assistance and album applications, but Google did not disclose the specific date, but said it will be released in the near future.
Google Assistant: finallyLogin iPhone
Since its release last year, Google Assistant has been an important part of the application and popularization of Google artificial intelligence in the mass market. Of course, this year's Google I/O gave Assistant a lot of drama.
According to official data provided by Google, the world has more than 100 million of the equipment used in Google Assistant, including a variety of intelligent mobile phone, equipped with Android watches and tablet computer system, of course, Google and Home.
This data is certainly very gratifying, however, Google does not meet: Google Assistant to landing iPhone, will support the installation of more than 9.1 of the iOS version of iPhone.
In addition to supporting iPhone, Google has also prepared many functional updates.
First of all, according to the official Google statistics, in the use of Google Assistant this year, there are 70% orders are issued by the natural language, so Google can increase the interactive function of text as a voice assistant, you can like a chat Google tell Assistant what to do.
Second, Google Assistant can finally schedule and set reminders. The schedule will be available today, but reminders will have to wait. For a voice assistant, the schedule and reminders are the last to be missed, and Google Assistant is really slow to shoot several times slower than Microsoft and Amazon's voice assistant.
In addition, users can now control more smart home devices via the Google Assistant. Google has worked with more than 70 smart home vendors, users can control their smart products through Google, Home and Android devices.
Google Assistant's voice commands are even stronger. Users can call directly on the Google Home via Google Assistant; connect to smart TV, and the user can also direct Google Assistant to play a particular music and video.
Third party service can access Google Assistant, Facebook Messenger Chatbot and the like, in keynote, Google demonstrated the application by voice assistant points out, it looks very good experience.
In addition, with the newly launched Google Lens, Google Assistant can help users identify the surrounding information. For example, when you see their love band played a poster, you can click on the Google Assistant Lens icon on the camera with posters, Google Assistant can help you get the band ticket information.
Google Home: no hardware updates, software upgrades, you can call
While the hardware base hasn't been upgraded, the new version of the Google Home has added a number of useful new features. The first is the "screen", this screen is to connect to Google Home on TV and mobile phone screen (or any display device), but not before Echo Show that comes with the Google screen strategy, called "Visual Responses" (visual feedback).
When you need screen interaction, you can talk to the Google Home speakers and show the results on your phone. If you want to watch TV, any TV that uses Chromecast connections can be a Google Assistant voice enabled device.
This means that you can not only let Google Home play your favorite music, but you can finally say the video you want to see. You can also see the weather and schedule on TV, take anything you want to do to push on the television, Google Home became the intelligent terminal to connect all the screen in the family, it is not like Echo Show that only a small compact screen, but a greater display space.
Another major upgrade is Hands-Free Calling, which is the long-awaited phone function, for example, when you say "Hey Google call Mom", it will directly from your phone book to your mother call. And speech recognition technology can identify who said this sentence, so the same sentence "Hey, Google, call, Mom" for different people, Google, Home will be heard later to different objects.
Obviously, the each function is updated in a challenge to the old rival Amazon Echo may, compared to Amazon Echo only in the United States, Britain and Germany to sell, this time Google Home in the UK, Canada, Australia, France, Germany and Japan and other countries will become an advantage, do not know Echo series then how to deal with?
Google Photos: make sharing smarter
Google Photos has been a private photo gallery for many people, bringing three new photos sharing this time to help users share the most meaningful moments of their lives to their relatives and friends.
The first one is Suggested Sharing, through machine learning techniques, Google Photos will help you automatically filter out the right picture according to the picture content suggest you who share, just click on the send, relatives and friends will be able to see your photos.
The second is Shared Libraries. For some photos, you may only share it with a particular person. The user can automatically share some photos with a particular friend, who will automatically receive them.
The third is Photo Books, Google Photos image recognition technology can bring you the best photos selected, automatically remove duplicate or poor quality, and make a photo album, you can print a solid album as a souvenir, priced at $9.99.
These features will be launched in the next few weeks.
YouTube: more users, more interaction
The stage is beyond all expectations YouTube products surprise came after CEO Susan Wojcicki, first with a dramatic video clips, shows wide range of YouTube and the influence of user groups, and then she said, YouTube riguan number has more than 1 billion people.
How can you play video sites that have such a broad user base? YouTube's answer is "interactivity" and "value-added services".
In order to further enhance the user interaction and video, YouTube joined the 360 degree video function, with a remote control, you can convert a 360 degree view in the video, you can see on TV, and different scenery, but also can watch the concert live a more feel personally on the scene.
The other is the service called Super Chat. As the YouTube video broadcast has grown 4 times in the past year, new "rewards" play has been added to the live video.
In short, you can take out real money to support your favorite anchor and get some more eye-catching marks as a privilege in your chat. In April this year on the "live" in the production of pregnant giraffe, giraffe donated money to more than thousands of dollars, and that our country pay by the way to buy gifts to anchor love is a reason, but the Super Chat above the "reward" is used.
Android O: for the next 1 billion users
In March of this year, Android surpassed Windows for the first time to become the largest operating system, but a lot of things can not be completed on the phone.
Android O will allow you to do more work on small screens. The picture in picture, you can seamlessly switch between two tasks, combined with the Android multi screen and multi task to optimize the deeper.
Using machine learning, double click text will automatically identify meaningful information, such as telephone, address, mail, click, you can directly access dial-up, maps and other applications.
To protect the security of Android devices, Google has launched Google Play Protect, built in every device that has installed Google Play. Play Protect automatically updates, detects and removes potentially harmful applications. At the same time, Android has also introduced the "find my device" function, which allows users to position, ring, lock and clear the device.
In order to allow developers to develop better applications, Android began to support the new programming language Kotlin, developers can greatly enhance the productivity.
The Android system covers from the entry to the various high-end equipment, in order to make the Android entry equipment at lower prices, Google proposed a new program called Android Go, the purpose is to let the equipment low configuration more capable of running Android.
Mainly through three aspects to achieve: Android systems, Google applications and Google Play stores. Starting with the latest Android O, you can run smoothly on the portal device. The application has been optimized specifically to reduce memory usage, such as YouTube, Go, Chrome, and Gboard. In one version of the Play store, these optimized applications are highlighted and all of the application directories are available.
The three initiatives will be carried out at the same time, starting in 2018.
AR: it's handy for indoor positioning
In June 2014, Google's AR platform, Tango (formerly known as Project Tango), was officially unveiled. With the camera and the processor, Tango can model the environment in real time, combine the real world with the virtual image, and realize various AR functions.
Earlier this year, Johnny Lee, head of the Tango platform Google's AR platform, said they had merged into the Google's VR division. Although this time he did not appear on the Google I/O stage, but some of the latest developments in Tango still worthy of our attention.
Google announced that Tango project will work with Google Map, launched an indoor positioning service VPS (visual positioning service), which can be in the absence of satellite signals indoors, but also precise positioning.
"This technology will come in handy in many places. "Google says outdoor positioning is still in the GPS world, but there's still plenty of room for room development in the VPS. The second generation Tango device, equipped with this technology, will meet with you this summer.
It is worth mentioning that, this I/O conference, Google also said it will spread Tango devices to campus, from this point of view, and earlier promotion of Cardboard is very similar.
VR:Google finally did VR one machine
Although the domestic VR market tends to be quiet by 2017, VR is still a very important strategic position for Google. First of all, Google VR responsible person Clay Bavor announced: Samsung, S8, S8, Plus and LG upcoming flagship machine will support the Google Daydream VR platform.
In the past, Google has been working on the VR box, and this time, Google has finally come up with an independent VR VR (standalone headset PC) that does not require a cell phone or an auxiliary operation. The two VR machines will be produced by partners HTC and Lenovo, and technical support comes from Qualcomm, which has already made great achievements in the field of VR as one machine.
The VR head of Google will significantly support inside-out location tracking (technology used on Tango), with a handle similar to that of Daydream View. In addition, no more specific parameter information is announced. Two pieces of equipment will be released at the end of this year. In October last year, carried out by Google "VR box" Daydream View officially released. Coupled with the release of the VR all-in-one, Google has taken a cautious step on the VR hardware.
Not long ago, Google announced the VR game studio Owlchemy Labs recruited into his command. Founded in 2001, Owlchemy Labs has made great contributions to VR content, and the games such as Job Simulator and Rick and Morty VR have gained high recognition from the industry and the players.
Rich hardware devices, expansion of the content team, Google Daydream's future performance, we look forward to.
Google.ai: show the best side of AI
Google today released a new Tensor processing unit of the second generation (TPU), which is a cloud computing hardware and software system, it will be on the line Google Cloud, called Cloud TPU, is mainly to improve machine learning processing needed a load, including training and reasoning, this can be said for the Google cloud the huge computing platform to bring the gospel.
Because of the integration of hardware and software in Google, TensorFlow has become one of the most advanced platforms for building AI software. The TPU is a dedicated chip designed for machine learning, and the first generation of TPU is used by AlphaGo artificial intelligence systems as the basis for its predictions and decisions.
This new TPU device provides up to 180 trillion floating point operations per second. However, in order to make these TPU more powerful, they can work better together. Google designed a customized high-speed network for each TPU that could combine to build a machine learning supercomputer called TPU pod. Each TPU pod contains 64 TPU of the second generation, the Google server has changed to 11500 trillion times per second (petaflops) super computer floating-point computation ability, so in order to accelerate to a single large machine learning model training.
These will be calculated in the future Google cloud engine on offer, and all related content on AI will be displayed in the new line of Google.ai, Pichai also said that "we want to make tens of thousands of developers are able to use machine learning and neural network through the neural network to further improve the AI he. "
One More Thing
At 2:52 in the morning, Google CEO Sundar Pichai came to power, after showing a video for you to introduce new features of Google for called "Jobs", which is the ability and the development of Google technology based on deep learning, when you open the Google want to find a job, in the search box to search your "work", you will see a new UI interface, in addition to see different work here will have different classification labels, to help you understand the different types of work.
It will not only through the label makes it easier for everyone to find their own ability to work, and accompany you on your search habits of learning, every time after you search, you will find the most suitable work to help the system will provide better suggestions in your search, of course, this everything must be based on the depth of learning technology Google. However, for such a better use of the search assistant, we are probably temporarily missed in the country.
Throughout the year I/O Keynote link, although not like Google Glass surprise black technology products, but for ordinary users, but there are a lot of new features, new technology is worth looking forward to, such as the Google Home in the Hands-Free call function of Google, Lens, and Smart will Reply into Gmail and so on. All this allows users who are accustomed to using Google services to have a whole new and intelligent experience that allows them to lead and experience other people's minds for our future intelligent lives.
At the same time, for developers, Kotlin become Android official language is undoubtedly a great surprise, when the news release, the scene broke out, perhaps the most enthusiastic applause and cheers today. In addition, regardless of the release of the second generation TPU chip, or the emergence of TensorFlow Lite, all developers dedicated to machine learning technology will benefit from it.
All shows that "steady" has become the main theme of this year's Google I/O, who is not too radical project and technology, but in the field of artificial intelligence such decisive, Google was steadily from the calculation platform to the application in all directions.
So the future of this industry is being staged in the town of Mountain View, not too far away.