You probably won't need to learn a foreign language, because a machine will help you communicate with foreigners; you won't have to grab the elevator all the time, because the elevator in your office stops at your floor at 12:03 p.m., and it knows that you've just stepped out of the office to get ready for lunch; you won't have to get your hair done to be annoyed that the new haircut doesn't really suit you, because an app on your cell phone will let you see your new haircut in advance, and it will let you see your new haircut in a new way, and it will let you see your new haircut in advance. See how well you match the new hairstyle ......
And from these applications end, we can clearly see that the development of the underlying computer technology is forming a new logic centered on natural user interface, big data, machine learning, cloud computing, and mobile technology: natural user interface makes the input and output of human and machine More simple; coupled with the increasing application of sensors will enable people to capture the amount of data through the machine is greatly enhanced, big data is therefore in urgent need. Big data will also lead to a situation where the computing power of local computers is clearly insufficient, and more integrated and powerful cloud computing becomes another key aspect. At the same time in the face of massive data what kind of data is useful, extract what kind of data seems to be more important, but these algorithms and arithmetic models are no longer set in advance by human beings, but by the computer based on machine intelligence automatically extracted and set. In addition, with the increasing popularity of mobile devices, mobile technology has become the link between local and cloud-based data interactions.
Here are some of the cool applications that Microsoft has released that will surprise people:
1, real-time voice translation
Imagine this scenario: you're alone in Helsinki, Finland, in a restaurant to order food, you said the Chinese language can be immediately translated into Finnish, and with their own voices to read out, and the pronunciation, tone and intonation of the locals and not the same. Isn't that amazing? I saw a similar scene at the launch, through this set of real-time voice translation system, a Chinese and English speaking foreigners to speak Chinese real-time communication, the machine in time to translate the language and happen, voice tone ebb and flow, voice recognition and translated sentences of grammatical accuracy is surprisingly high.
The more simple and comfortable the application looks the more complex the technology behind it. There are a number of technologies involved in a short three-minute conversation. The first is that with deep neural network (DNN) technology and simulation of human brain behavior, Microsoft Research has developed a speech recognizer with much higher discriminatory power than traditional technology. The second is personalized speech synthesis that retains the user's individual voice characteristics. Again, there is the continuous accumulation and development of machine translation. In the translation scene shown on the scene, the translated sentences are almost free of word and grammatical errors, sounding authentic and smooth.
We look forward to the day when machines can help humans break down barriers between different languages and cultures, and people will finally be able to communicate smoothly with anyone in the world. Who will be the "victims" of this technology? The education industry. With such a machine, it seems that we don't need to learn English anymore. Hong Xiaowen expects the technology to be officially launched in 3-5 years.
2. Kinect-based sign language translation system
Imagine another scenario where a deaf mute asks a stranger on the road for directions in sign language, and the good Samaritan, even if he doesn't know how to sign, can understand his words without much trouble. Through the intermediary of computers, the day is also coming when deaf people are realizing their new hope of having a "voice" through machines.
The technical logic behind this is that the computer uses a camera to capture the deaf person's sign language movements, recognizes them as corresponding words, organizes the utterances and then vocalizes them. The system supports the recognition and translation of both Chinese and English sign language, and can reverse translate text messages into sign language.
The core technology is the Natural User Interface (NUI). By natural user interface we mean that human-computer communication is realized in a more free and natural way, and that the computer's input system (not the input method, mind you) becomes more intuitive and easier. The earliest input systems for computers were keyboards and mice, in recent years the computer's input and recognition of speech has been realized, and the next step will be somatosensory input, which turns on the computer's webcam and you will be able to manipulate the computer through gestures, facial expressions, and even body language. The technology, called Kinnect, which was once used in Microsoft's somatosensory gaming console Xbox, is starting to become more refined and sensitive, recognizing not only large body movements, but also subtle changes in finger movements when a deaf person uses sign language. This will not only make it easier to control computers, it will also allow computers to be used in many new applications.
And in addition to Kinect For Windows, this includes technologies such as Bing Dictionary, Microsoft Translate, language modeling, and 3D modeling.