Recently, the assessment report of Companies and Markets said that the global speech recognition market will become more diversified in the future, and the accuracy of software will be greatly improved. It is expected that by 2017, this market will reach $133 billion. Friends who have seen the movie "Her" will be impressed by the plot of speech recognition and control. We have previously introduced the difficulties in the implementation of this technology, but in the future, what are the scenes of speech recognition technology applied to? When it comes to speech recognition, the easiest thing we can think of is Siri, which doesn't tell jokes. Siri's technology comes from Nuance, the world's first publicly available speech recognition company. A few weeks ago there were rumors that Baidu wanted to acquire Nuance to help its voice assistant application, but this matter has been denied by Yu Kai, director of Baidu Deep Learning Lab (IDL). In November last year, it was reported that Samsung had negotiated with Nuance about the acquisition, but was shocked by the fact that it had been losing money for six consecutive months. On the other hand, Apple does not want to rely on Nuance technology for a series of self-organizing activities, which also prompted Nuance to find new exports. As a business that is difficult to bring in cash flow directly, voice recognition has made Nuance shattered. Although the overall situation of the company is not optimistic, we still see Nuance's attempts in commercialization in different fields. Medical field Some time ago, Nuance announced on the official blog that he would report on the medical field. Not only is it simple to track the movements and heart rate through smart watches, but also to match the appropriate services, such as the right restaurant or food, depending on the person's physical condition. Of course, these are mostly based on wearable devices. In addition, they also considered more scenarios, such as emergency voice help, doctor-patient dialogue archives, call center dictation, and so on. Due to the strong evolution of the professional vocabulary in the medical field, as long as a complete database is established, the disease name and drug name can be relatively accurately identified. Existing domestic applications, such as medical records and Zhenli, provide doctors with a cloud space for secure storage of medical records, which is convenient for finding cases and supporting voice search. Smart car There has been a lot of attention on the issue of driving safety. Last year, someone designed a car screen that could solve the problem of excessive distraction of the driver's operation of the touch screen by using simple gestures with multiple fingers. In fact, Nuance has already had a custom car-level voice platform, Dragon Drive, and won the 2015 CES Innovation Award. By connecting the vehicle platform to the mobile phone, Nuance can help users achieve voice control GPS navigation, information transmission and reception, telephone access, social network updates and so on. Smart wearable The winter in Beijing is getting colder and colder, and there are countless times when you need to use your mobile phone to live and die. Calling, sending WeChat, checking the route, calling the car, if you can use a wearable device to achieve through voice control, that would be great. At present, I am going to ask the team to make Ticwear, a smart watch operating system equipped with Moto 360, which can realize direct voice control drip taxi. A year ago, Intel developed the prototype of wearable device Jarvis, which can be worn on the ear and then connected to the mobile phone. In order to understand the instructions issued by the owner and give voice feedback, it acts as a personal voice assistant role like Siri. . This is not the scene in "Her"! Smart home Before we introduced a smart bed cover called Luna, we wanted to collect more human data in a new way, and integrate various smart homes. Speech recognition is perhaps the best way to become an integrated portal, especially when we are lying in bed and don't want to see any screen at all. What's more, Nuance has long developed technology that can control TV and mobile phones. 36 氪 author Da Feige once sent a circle of friends, "Who can use the Internet thinking to transform the family bathroom, it is very troublesome to pick up a speaker every time you go in. Moreover, cutting a song, taking a temporary call is simply anti-human." Although the smart home bathroom has not been heard yet, there are already smart speakers that can be controlled by voice, such as the Xiaozhi super speaker we introduced. education field One of the difficulties we have mentioned before speech recognition is the differentiation of many languages ​​and dialects. On the contrary, it can be used in the field of education. Allowing the database to provide a standard can provide a more convenient and efficient way for oral assessment. In terms of assisted learning, I think of the 36-day open-day Shanghai station a year ago, "Zidong Interpretation", a cloud-based Chinese-English simultaneous interpretation application. It integrates speech translation, speech recognition and speech synthesis technology. You only need to speak the phone to the sentence you want to translate, you can get the Chinese and English double voice broadcast results and visual text results. It supports continuous voice input and even automatically adds punctuation to your tone based on your tone. All in all, speaking is the most natural way of communication for human beings. We can say that artificial intelligence is pseudo-intelligence, and machines never live like humans, but machines can become more and more understanding. Shandong Longze Mechanical Equipment Co.,Ltd , https://www.pelletmachinefactory.com