This year's I/O Google conference first show is to come to an end, followed by a small series found that in fact this session of the I/O Conference on various systems and new features are separated from a word - robot learning system. From the core of the Google assistant to a variety of operating systems and even all the new interactive way, Google has mentioned this program. Next, we will summarize some of the situation of this event.
When it comes to the I/O Google conference, the mobile phone is most concerned about the Android system is probably updated. From yesterday's perspective, the small series has been mentioned, this time the Android system update more inclined to enhance the system experience, in terms of practical innovation is not much bright spot.
Next to a rough look at the main N Android system update bar:
Performance: API for new, API Vulkan upgrade processing ability to reduce the GPU requirements; to join the JIT compiler, App installation speed upgrade, take up memory to reduce, battery life upgrade.
Security: Join file encryption, seamless system updates, new media framework.
Notification: notification bar interaction improvement, can be directly in the notification bar reply application information.
Function: the split screen function, pip.
In addition to the Android system, Wear Android has also been upgraded, the following is the main feature of the Wear Android 2 update:
Interaction: improved the operation of smart watches, improved on the slide and slide gesture and function. Watch based on Wear Android 2 system can rely on Wi-Fi or LTE to complete the data transmission.
Notice: according to the different characteristics of the information to adjust the background color of the notification, the information will gradually fade, still keep the operation of the pop-up notification.
Input method: join the robot learning system as the bottom of the input method, provide intelligent reply, handwriting recognition, and a complete set of English keyboard.
Movement: join the movement to identify a new API, more accurate to identify the state of motion
The first VR platform: platform and application of intelligent robot into a big bright spot
In platform and application, follow up the VR pace this year, Google launched a daydream of VR platform, Google's VR platform daydream to mobile phones as the core of the system, but on top of this on the phone has a more requirements: system, software, sensor, indispensable, performance, low latency, VR system UI, these improvements are Android n of the core part.And at the meeting, Google demonstrated the VR Daydream platform system operating interface.
And this time the main highlights, probably still in the application. Google has spent a lot of time on the interpretation of the application, in fact, it is Home Allo, Google and Duo. Three applications are based on Google search and Google assistant, and the new Google assistant based on robot learning system. Three applications on the basis of Google search and Google assistant to achieve real-time interaction between people and applications. For example, Google home in the Google search for speech recognition based on smart home real-time management and ordering all rely on network, like taxi service life can be achieved. Has the basis of voice control, Home Google and even do not have to open the phone App, you can achieve some of the smart home functions. Allo and Duo are more inclined to help people communicate with the robot, which will be mentioned in the next robot, the trouble to look down.
Ultimate Boss: robot learning system into a leading role
Three key applications have a lot of robot learning system, then the robot learning system is what? First, this must be referred to the Go (Alpha), the main working principle of the AlphaGo is the depth of learning". In simple terms, as long as the accumulated experience, the robot learning system will become more and more smart, more and more understanding of the user's idea. The experience is mainly derived from the data, just, Google has a leading global advantage in the data, which makes the robot learning system has become more powerful.
Google, however, is very clever to put the system into a platform for human assistance. So, at present, as an auxiliary robot learning system will make the interaction become more simple, will make your hand machine more understand you, but it is still a machine.
Google launched from the three applications to explain the current Google robot learning system. Google assistant is a product of the robot learning system, by this system, Google assistant will use the huge Google search for your information given reaction. As an intelligent robot, Google's powerful place in and - it is able to identify the problems you are given, but also for your further questions to verify the problem, to find more suitable for you. On this basis, it can also be based on your daily operation to learn, the result is to use the number of times and the ability to enhance the recognition. Allo and duo two applications are based on Google's assistant, but Google will Google assistant's ability to upgrade again, in allo and duo, assistant to the Google will be able to identify friend send to your information, you can also in need to give your friend a reply to avoid busy ignored the embarrassment of friends.
Visible, this session of the I/O Google general assembly, compared to those systems and platforms, the main update is still in human-computer interaction and robotics learning system. After Go Alpha, Google will be applied to the system of this system to the level of life. Big data is still a very powerful Google, big data with the depth of learning the robot learning system may be in the future to provide a stronger auxiliary function.