Can human-computer interaction rely on voice commands only?

Absrtact: In the last article when we discuss smart home, don't forget the robot, I mentioned that anthropomorphic AI may become the center of intelligent home equipment. So the question is, can human-computer interaction only rely on voice commands? The turn of the machine to the environment

In the previous article, "When we talk about smart homes, don't forget robots," I mentioned that anthropomorphic AI could be a central device for smart homes. So the question is, can human-computer interaction only rely on voice commands? What about the interaction between the machine and the environment?

In the fall of 2013 The Intel Information Technology Summit (Developer FORUM,IDF). Intel has brought a lot of products to the Intel chip, which is very eye-catching, that is the prototype from the robot of the Walle, her name is the E-Wally. The most important function is: first, greet people, and shake hands; third, Intel placed a screen in the robot's abdomen, allowing people to see how the Robot "sees" or perceives the world.

Look at "tile power" that Meng Meng Big eyes! This is the best gift Pixar has given humanity. But I'm sorry, from a technical standpoint, these two piercing eyeballs are just decorations. Robot chest Microsoft Kinect 1 is the so-called "eyes" and "ears." The robot uses Kinect to get images and depth information to complete the process from identifying people to greeting and shaking hands. Because Kinect is also equipped with a high-definition microphone, so incidentally "listen" also resolved. In addition, the Kinect itself has a certain computational power, also played a part of the brain role.

Do you understand? It is through the depth of the camera to get the real world information, and transform it into the virtual world of the data needed, and then processed. Now there is a fashionable word to describe this whole process, called "machine perception", perhaps immediately the word will be like "cloud" even aunt can say. In short, you can understand that the depth of the camera is the next generation of "computers" such as the most important organ of the robot, this organ will assume the responsibility of a new generation of human-computer interaction-from people and mobile phones to people and robots.

In my last article, I introduced the Spider-Type robot, and the little guy was carrying a webcam from Asus ' Xtion. With Kinect 1 as a half-parent relationship, its core technology comes from a father's Dna--primesense, a company that has long dedicated to the deep camera civil. PrimeSense is not the ancestor of the depth camera, but this new human-computer interaction revolution, but this is the first shot, in the primesense behind, is countless giants for its torch. In simple terms, PrimeSense solves a number of mathematical algorithms that help them design chips that can quickly process depth information, which is licensed to the hardware manufacturer, which is the core component of the depth camera. The popular metaphor is intel in the image graphics world.

PrimeSense was founded in 2005, in 2006 announced the completion of the 3D induction chip research and development, in the year's E3 exhibition with Microsoft to establish a link, and catalytic code for Project Natal Mystery Project. By the time the 2009 E3 exhibition, Microsoft had released the Kinect 1, which was licensed for the PrimeSense chip. It is also the origin of the PrimeSense delegation, including ASUS. But it is rumored that Microsoft abandoned PrimeSense in the Kinect 2 o'clock for failing to acquire PrimeSense. Finally, on November 25, 2013, Apple finally finished 360 million of PrimeSense's acquisitions.

You see that? From Microsoft, Intel to Asus, and ultimately Apple, the Giants in the computer industry have been linked to the younger company. Calcalist, the first Israeli media to buy PrimeSense, described Apple's intentions: Apple acquired key technology in the area of indoor entertainment by acquiring PrimeSense.

After this, countless media are predicting Apple will be in the living room of the battlefield with Microsoft, Sony break off wrist, from a new generation of Apple to ITV. But the acquisition is now 1 years old, Apple seems to have no news in this regard, how much is a bit disappointing. But it also fits Apple's style: Secret development and Blockbuster.

You have to think that the industry will be primesense by the disappearance of insipid, then you are wrong. In fact, from the beginning of the acquisition of PrimeSense was formally opened a new human-computer Interaction War Prologue, once in touch screen or the mouse era of many giants, are in the rapid admission after the layout. Because PrimeSense is bought by Apple, its foreign-mandated technology will expire in 2015, so those who use PrimeSense technology have to look for alternatives. Many people see new ways to make money, such as many domestic start-ups, such as millet. Others see the point of subversion, hoping to recover in the mobile era, such as Intel. Others are afraid of subversion, like apples.

Some of the names of companies from China may not be familiar to you, but they may be the next generation of stars. Next, I will be divided into two parts, domestic and foreign, to resolve the layout of the Giants or would-be giants:

Domestic start-up companies:

Glitch Deep pupil main intelligent security company, prior to 36 Krypton had glitch deep pupil has reported, they also claimed to have built the security equipment with the depth camera, the Microsoft Incubator Phase II company, see the team history is very dazzling. But I judged their actual competitiveness should be on the software and storage side, on the hardware side will generally tend to purchase more cost-effective solution.

Obi's core goal is to replace the PrimeSense, they from the day of the establishment is to be able to make up for PrimeSense no longer authorized, the domestic left market blank. Their goal is to provide equipment for new human-computer interaction based on Android. One of the main markets is the game market, such as those featured in the game's Android box, especially in the case of video TV. le-Vision TV has already had the function of the body feeling game, they choose precisely is the xtion that the PrimeSense authorizes to produce. When Xtion inventory is not available, Obi products can be Plug and play. It is reported that they have completed the core chip of the shape of the film this October, is expected to be in February 2015 (probably during the spring Festival) to disclose more information.

Beijing ice fruit This company from the business registration information, mainly engaged in technology import and export. From multiple sources, they are commissioned by the XI ' an Jiaotong University in a research institute in the Kinect 1. Like Obi, their goal is to replace PrimeSense's remaining markets after the cancellation of the authorization. The company's personnel composition is not yet familiar, if you know the reader, you can contact me by mail.

Cool Sense Technology parent company is digital video, stock code 300079, the main shareholder of Tsinghua Science and Technology Park, belong to Tsinghua's school enterprises. Tsinghua's graphic image laboratory has a very important position in computer vision and can be used to speculate on the company's technical strength. They declared that they had completed the research and development, and the goal was to replace PrimeSense.

Foreign computer vision must be the first to talk about the famous Microsoft Kinect, but their deep camera core technology is still from the authorization, Kinect 1 is from PrimeSense, Kinect 2 from Canesta. Microsoft's long-standing goal is to apply a depth camera to the gaming scene. But there are two things that I can be sure that Microsoft will start miniaturization of the Kinect and apply it to new scenarios.

First, with the resurgence of robotic vision, Microsoft's previous Open-source robot, IROS2014 Chicago, annotated the Kinect-enabled core sensor, and then recalled that Bill Gates first talked about 3D sensors at the D5 summit in 2007, At that point they have already cooperated with PrimeSense, so gates will say they already have the goods. The vision that Gates talked about at the D5 summit was the ubiquitous camera, where people would go, and there was an operating system, a robot that could go anywhere through the Kinect, or where it could be followed. The goal of Windows 10 is to be a hub for the internet of things so ubiquitous that the Kinect or other "organs" with more powerful functions will need to be deployed everywhere.

It can be concluded that Microsoft has been able to maintain its technical advantage in this area for a certain period of time, as the Kinect 2 is a leap-forward innovation product, compared to the Kinect 1, which has been freed from PrimeSense. In addition to Microsoft, large companies and foreign startups in this area are:

After PrimeSense was bought by Apple, PrimeSense announced that it would stop the authorization in 2015, and it appears to be the heart to make the hardware that contains the depth of the camera. TV or cooperation, or an apple robot? The author no longer guesses. But here's the interesting fact that Pirmesense has done a miniature depth-camera development earlier than any family. In the second half of 2012, PrimeSense completed the development of a miniature depth-camera device code-named "Capri", which targets laptops, tablets and even mobile phones. In the near future, it may be seen in the iphone.

Pebbles interfaces The company was invested in millet, because Millet has no relevant technical reserves, the future will not have this plan. Through this acquisition, Millet has been given the relevant authorization for the 3D Sensor (core chip). Also signed a three-month exclusivity agreement, I believe that in the future Millet smart TV and box will see the use of body induction.

Intel purchased Softkinetic's core technology and then launched its own deep camera device--realsense. RealSense has a corresponding model for Android and Windows, as well as an ultra-polar and flat version. As you can see, Intel is trying to make the depth of the camera smaller, and iron heart to seize the next generation of human-computer interaction portal. Like millet investment pebbles interfaces and PrimeSense, softkinetic from Belgium.

Google buys the technology from Movidius, which is mainly used in Project Tango, the concept of 3D sensor phone.

Leap Motion The company has long been known to the people, before they also cooperated with Shenzhou entered China, in addition to the main computer-side man-machine interaction, but also want to enter the car network such areas. The accuracy of operations has always been the focus of leap Motion, but it may have caused the leap Motion to experience poor performance due to excessive emphasis on accuracy. In addition, they are dealing with the relationship between the developers repeatedly problems, resulting in weakness.

If there is a chance, I should make a detailed introduction to PrimeSense, pebbles interfaces and Softkinectic, a personal father from Israel. But the omission here is intentional, because in fact these companies ' research and development is very expensive, which is why PrimeSense in 2012, the company's employees from 190 to 50 people. Without Microsoft's support, they are actually losing a steady stream of revenue, while Godzilla-level Microsoft has the ability to set up a new portal and direct money to burn a Kinect 2.

It takes a lot of money to complete research and development, which is the reality of the depth camera upstream. This is why only Microsoft, Apple, Google can be called the real players, and Leap Motion such vendors are often weak successor. Pebbles interfaces is directly to China to complete the financing, standing behind them is millet such a new giant for its delivery shells.

Perhaps the admission of the Super Giants heralds the advent of the next generation of human-computer interaction, but the collective layout of the new Giants is also indicative of the high threshold for the industry in hardware. So what can they do to people who are doing business in this field?

Do smart home, home robot, you can purchase equipment to the above companies, the interactive requirements are not high can choose the domestic company's program. Used to do the Internet to do the app company, as with the author, honest to do the application. However, in the application of new human-computer interaction, the understanding of each scenario will rise to several levels of requirements, even developers have to read more data from the platform transfer.

Because in this era of ubiquitous sensors, we will be in the real world cheap to capture a lot of data, read the data, to get more help from the virtual world.

Are the Pirates ready for the revolution?