"Voice interaction" breaks through "button"
In daily life, the button is almost the most common man-machine interaction. In addition to the computer keyboard, the operation of the copier, automatic washing machine, microwave, mobile phone, etc. in the office is performed by the button. Although the copier has become more intelligent, people have received more information through its LCD screen, and touch screen control is also more convenient. But "lazy" users are still in questioning: Can you do any button? Don't you direct the "it" directly?
In fact, there have been dozens of mobile phones with simple voice dialing, but "should we wait for ten years again?" - Calculation of the 2003 "Branch of the Microsoft Asian Institute "At the General Assembly, Li Kai Fu said in the speech:" The human inventive language is to communicate through the language, people also want the machine to communicate with themselves with themselves. "The voice became the most optimistic man-machine interaction in the future.
Voice has more advantages over other interactions, the earliest voice technology plans to include speech recognition, natural language understanding, and speech synthesis due to the "Automatic Translation Telephone" program. The research work of speech recognition can be traced back to the Audry system of the AT & T Bel Laboratory of the 1950s, and the researchers have gradually broken through the three major obstacles of the big words, continuous voice and non-specific people. In the early 1990s, many famous large companies such as IBM, Apple, AT & T and NTT were investigated by the practical research of speech recognition system.
At present, there have been some commercial products at home and abroad, which is relatively representative of IBM's VIA Voice and Dragon's Naturally Speaking. In September 1997 IBM first launched a VIAVoice Chinese continuous speech recognition system in China, spending a lot of funds and cultivating the market, so that everyone knows what is voice technology. IBM is still constantly introducing new ViaVoice versions, applying voice technology to PDA, smart cars. It also provides voice development tool SDK, hoping to create a full range of voice platforms. Microsoft also integrates speech recognition technology to multiple leader products, including Office and Windows XP, the latest speech recognition server software SPEECH Server is ready to be released in the first half of 2004. The software allows users to operate computers using voice commands, and companies can also use it to create a service similar to the automatic telephone system. These are the categories of nick identification. According to Dr. Liu Qingfeng, General Manager of Bo Qingfei Company, speech recognition has reached 99% accuracy in identification, a "sesame opening" opened the door is not legend, but speech recognition Basically, "fun", cannot be used in high confidentiality, such as banks. "
The voice recognition allows the computer with "ear", but the brain like human human beings is still a dream. So far, "Automatic translation call" is the initial plan is still the goal of researchers.
Let the computer need to use voice synthesis technology, which is a text to speech. In the country, BSFA, Microsoft Asia Research Institute, IBM China Research Center, etc. have excellent applications in this area. Dr. Liu Qingfeng believes: "After the PC era, the screen is getting smaller and smaller, and it is a must to listen to the mobile terminal is an inevitable requirement. If there is no terminal display screen, voice technology must be used. Voice synthesis has reached the industrial application stage In the field of social information services, mobile terminals, providing information services and consultation have been completely no problem. "Not long ago, the Boardian News flew to design a speech inquiry system for Beijing 2008 Olympic Games, so that the tourists from all over the country will be A unified number query Olympic information and Beijing city tourism, shopping, dining and other information. This system is expected to be basically completed in 2005, providing multi-language, multi-dialect, multi-spokesperson, multilingual tone, and has a personalized voice service with self-learning ability.
"Handwriting input" beyond "voice"
The owner called the robot to cover the child to cover the child, but the robot has covered the "cup" to the child, and finally the owner wrote "quilt" on the mobile phone to pass the robot. The robot understands the meaning of the owner. This is a scene in Nokia's mobile advertising. To implement voice "interaction", the machine must understand the identified language. Experts in this area predict that human beings are still unable to communicate with the natural language, mainly because the natural language cannot be fully expressed in a limited set of expression, and the intention of understanding the speaker from the statement is more difficult.
The handwriting input is calculated in addition to a traditional input method, and the design of the handwritten pen on the market fully meets the requirements of practical applications in the identification rate and identification speed. Several larger manufacturers can reach 99%. The handheld computer is a successful model for handwriting input technology applications. On the one hand, the movement allows the use of the keyboard to be less and less; on the other hand, in WIMP has already changed today, today, paper, pen or well preserved in everyday life. How to make these contents of the paper into part of E-mail? People began to study handwritten input, digital ink and other technologies.
The handwritten pen using a compactive technology can capture the changes in handwriting pressure, and there is a strong sensitivity on the thickness of the stroke, which is the mainstream in the market. For example, Wacom technology introduced by manufacturers such as Mengyi, improves the accuracy and input speed of product identification through the upper level of pressure. IBM China Research Center has been committed to developing continuous handwriting recognition techniques, people can directly write and paint directly on a notebook notebook in a normal paper, and the page writes automatically saved in the memory of the device on the notebook floor. This technology is also applied to multiple products of IBM, including IBM ThinkScribe, IBM Thinkpad Transn. Microsoft Asian Research Institute has made great progress in digital ink technology, which enables people to enjoy the convenience and benefits of handwriting without handwriting, and have become one of the core technologies of a tablet.
In December 2003, people saw in Hanwang Technology Handwriting Input Technology Application Provenance, the application of handwriting input technology has far exceeded the traditional computer input category, including tablet, intelligent fixed call, various palm information In the field of terminal, touch screen, POS machine, smart refrigerator, especially in the application of a new generation of smartphones, has already shown outbreak. In the computer output, the main development direction is to make the user more intuitive and convenient access, so the display is still absolute mainstream. When mobile devices have "slim", the size of its display is still dominated by user, presenting a gradual trend. At the same time, virtual reality technology is gradually widely used to present natural images for users, and holographic techniques are used instead of stereoscopic displays or stereo glasses. Maybe not a long time, the virtual display of the "Hacker Empire" will become the latest human-machine communication.
"Mouse", "Eye Market", "Brain Signs"
Interaction basically inseparable from the user's visual, then if the user can use the eye (directly using the brain thinking "brain syllable" is still too far away) to manipulate the graphical interface is not more convenient? "Objective" came into being.
In 2002, the research team of the London Imperial College took several years to study the relationship between human eye activity and brain perception, and finally invented an infrared sensing device specially used to capture the fine action of eyeballs, which is called "eye label" . It can identify the movement and gaze of the eye on the display. If the computer user pearss a link icon on the screen for more than 1 second, the computer will automatically open the link; if the user focuses on the icon of the print or store, the computer The corresponding operation is automatically performed. However, the inherent jitter exists during his eye movement and the data caused by the eyes of the eyes, obviously interferes with the signal. In addition, if the mouse cursor always moves with the user's sight, it may cause a dislike because the user "casually" looks not always "means" what. Ideally, the user wants to issue a control when the system processes its sight input, and in the contrary, the movement of its sight is ignored. Let the computer understand this "hope" of this user is not much likely. However, hope is long, because researchers believe that human use of eye search and stare at a target, far exceeds the mouse with hand. How to make the computer "see", belong to the scope of computer visual research. Many of them have been applied to actual identity authentication techniques, such as iris identification, face recognition, all by collecting images and results. Existing technologies can be photographed through the camera, and then use image analysis and identification technology to identify.
It is not a problem with the image "see", and a small camera is competent. The key is how to get the required information from the image. To make the computer like a person "see", the most fundamental problem is to computers to the ultimate analysis and understanding of the image, although the shapes, colors, textures in the picture have provided more information, but can be used A priori knowledge is thinking, identifying, and returning to the old problem of manual intelligence.
At the same time, people start research to let the computer go to "see" how people control the graphical interface. The development of sensors has made a variety of interaction techniques, such as data gloves for identifying gestures. It detects more complex hand movements, including the position and direction of the hand, finger curvature, and classify according to this information. Similarly, Simgraphics Developed Virtual Actor System (Vactor) developed by the user with a helmet of a sensor that touches the face of different parts to control the computer generated an expression.
"Multi-channel interaction" business prospects
Despite the high voice input efficiency, you can't imagine that every employee is in the office, running your computer. Although each interaction technology itself has developed to a certain level, there are also flaws that cannot be abandoned in the short term. In the case of unsuccessful technology, relatively complex interaction is inevitably complained by the user.
And these interaction techniques can only be used as a means of input or output, which itself cannot be "interactive", only to apply them to the interface - the concept of multi-channel human-machine interface is intended to complete this integration. There are two purposes to be achieved by multi-channel technology itself: one is nature of human machinery, so that people can use more natural, efficient ways at any time, and anywhere in the Internet. The exchange of communication; another purpose is to make a lot of technologies more perfect, such as let voice recognition and handwriting is a good concept.
"There are two different starting points for new interactive systems: one is to let more users can enjoy the benefits of computing technology; the other is to make existing computing devices more convenient to use. From The improvement of the base of the desktop, such as the latter, such as a wireless mouse, is the latter. "Dr. Wang Jian, the Director of the Multi-channel User Interface Group of the Microsoft Asian Institute, said" "Existing interactive modification is not very large."
In the past five years, all kinds of human machine interactions include multi-channel human-machine interfaces that have spawned a new generation of human-machine interactions, many people (including users and merchants) are looking forward to the birth of "making mouse can't exist" . The transition of the mass change is so unpredictable. How many years can be achieved after "next-generation human machine interaction" is also difficult to predict. However, once it appears, it will become a decisive factor that the computer is completely moved to the public. Unfortunately, it is not only possible to accurately define the next generation of human-computer interaction, where is the future "killer application", now it is still not known, but the business opportunities and commercial strength of human machine interaction technology, but have been Many business meet. Not only in the computer industry, network, mobile communication, information appliances and other major industries, but also in the palm of Mark. Like Bluetooth and 3G, only a carrier of human-computer interactive mode, it is stunned by the rivers and lakes and the market disputes. However, Dr. Liu Qingfeng, the Department of Commerce, talked about the actual application of the pressure as the technical study when talking about the business prospects of human-computer interaction. "Consumers can't see the practical application prospects of a technology, which requires vendors to promote. However, due to the restrictions of research funds, considering investment output ratio, fresh enterprises are willing to do market wastewors like this year. And it is affected by piracy, and the company is careful when making market development. "
Classic scene in human machine interaction
The oldest human machine interaction - perforated card
The world's first digital computer ENIAC uses an external program that enters the input and output by reading the cardifier, tacker, and the typewriter. Piercing card should be an earliest human machine interaction, inspiration from the 1801 French textile engineer Jamada invented jacquard, the machine controls the jacquard operation steps with a small hole on the perforated paper. At the end of the 19th century, the US Statistics H.hollerith Dr. H.hollerith is borrowed from such perforated cards to complete the large-scale data processing of the whole American census, although his original intention is not in computer interaction. This input data using perforated cards is used to be used in the 1970s.
Breakthrough mouse
This is a familiar allusion. Douglas Englebart produced the first mouse device in California - only one button, the shell was carefully engraved with wood, with a metal roller at the bottom, and did not be Pay attention. Until December 1979, the Scientist of the Xerox PRAC Research Center demonstrated the window, icon, menu, and the "mouse" cursor, the major historical significance of the mouse is at this moment. Really get displayed.
Interactive manipulation of "virtual elegance"
On December 5, 2003, a demonstration in Illinois Lake Forest formed a "virtual" screen, and the user moves by touching the image with the image, select the icon and does not require Dai special-made gloves. The 29-year-old graduate student named Dyner invented this technology called HeliodisPlay, using electronics and thermodynamics to change the properties of the air entering the machine, and the air injected will generate an image. Laser Tracking System Tracking User Finger's Effectiveness Control for images of "virtual elegant". Although it sounds around, it is difficult to match the mouse, but it leaves the binding of the desktop to make the user more naturally manipulate the graphical interface in a particular occasion.
