Baidu building deep learning lab in silicon valley

Beijing time on April 13 news, according to wired magazine reported that baidu has build laboratories in Cupertino, California.

in January, there was news that baidu will create “deep learning (deep learning) laboratory, deep learning is an emerging computer science, mainly is to use hardware and software to simulate the human brain. Now it seems that baidu set up operations center in silicon valley, lab, not far from the apple team will also provide support in China. It’s just hired top researchers in Cupertino, will hire more employees by the end of the year.

baidu voice recognition team director Kai Yu introduced in silicon valley deep learning

baidu will lab named “The Institute of Deep Learning (Deep Learning research center, IDL)”. Like Google, apple and other companies, baidu is trying to explore like people learn computer system. Baidu’s voice recognition team director Kai Yu recently arrived in Cupertino, hired top researchers. Kai Yu said: “with deep learning to imitate the function of the human brain, ability, intelligence, we have a big dream in this respect. We in progress every day.”

if you want to compete with Google, build a lab in its backyard is meaningful. Kai Yu said: “in silicon valley, we can contact with huge talent pool, here are really top engineers and scientists, Google has the advantage.” About a year ago, baidu set up the first office in Cupertino, introducing different other employees, and then to the “deep learning” research taken a big step.

in the 90 s and 2000 s when the deep learning research fell to the bottom. Artificial intelligence research advance the system of processing huge amounts of data, rather than to build “neural network” to imitate the brain subtle activities. Google’s search engine is around “deep learning” copy a shortcut to drive cars also USES a similar method. Now, “deep learning” research, Google is also promote the development of it.

recently, Google has hired “deep learning” as the father of Jeffrey Hilton (Geoffrey Hinton), IBM has gone very far in brain computer model research. Apple iPhone Siri voice recognition system USES the technology of “deep learning”.

deep learning expert at the university of New York Yann LeCun thinks, baidu is a little bold to set up a special “deep learning” laboratory technology has a long way to go. Through the establishment of laboratory, however, Yann LeCun think baidu can attract top talent and creative engineer, looking for all kinds of innovation, such as similar Google Glass type of innovation.

in fact, Kai Yu, one of the research project named “Baidu Eye”, many people think that it is in accordance with Google’s project Glass. Kai Yu said that so far, the main task of the laboratory is to explore deep learning algorithms. He said: “we want to focus on a few.”

in November last year, baidu launched the first voice search service, based on the deep learning it says tool error can be reduced by 30%. The improvement of the technology may change the way people interact with technology.

at present, the link to search for a bit more. Kai Yu said: “the user to read all the web page to know the meaning of it.” He hope to change this kind of operation form. Kai Yu said: “we want to fundamentally change the architecture of the whole system.” Also is to build the new algorithm, it can identify the pictures, understand natural language, understand the relationship between online all elements, find the really want to find east west for the user. Need to like people, in other words, the operation of the algorithm, is faster than people.

translation: sohu IT