In addition, he also confirmed the news that Baidu will expand the scale of the Silicon Valley Institute. The new office area is less than one mile from the current Baidu American Institute.
"The number of people will increase about twice as much on the current basis," Coates said. At present, Baidu has about 200 employees in the American Institute.
(Baidu Silicon Valley Director of Artificial Intelligence Adam Coates gave a speech at EmTech Digital Conference)
Last week, the sudden departure of the chief scientist of Baidu Wu Enda caused new questions about the future development of Baidu artificial intelligence research. Baidu announced that the president, Zhang Yaqin, will be the chairman of the Baidu American Research Institute and be responsible for Baidu's key personnel attraction in the US R&D center, brand building, and strategic coordination between the US research and the domestic business.
Adam Coates said that the nature of his work at Baidu will not change and he will continue his previous research on artificial intelligence. Prior to joining Baidu, he followed Wu Enda's postdoctoral research in artificial intelligence at Stanford University.
"When I was on a business trip to China, the best thing that gave me was that everything changed very quickly," Coates said. "It's a completely different rhythm from Silicon Valley."
He said that when Baidu engaged in the background of artificial intelligence research, for many developing countries, the development of mobile Internet paved the way for the era of artificial intelligence.
"If you connect to the Internet for the first time in 2017, you're probably connecting from a mobile device," Coates said.
He believes that it is based on people's changes in the way the Internet is connected, and the way human-computer interaction will change at the same time. Mobile and IoT will become interactive voice and artificial intelligence.
"In the era of mobile Internet, we need a new human-computer interface," Coates said. "And voice is what we think can affect 100 million people."
Baidu Silicon Valley Research Institute has developed Deep Speech technology in the past few years. Unlike traditional speech recognition systems, Deep Speech fully uses deep learning technology to train hundreds of hours of speech data to realize speech recognition.
Earlier, Wu Enda once introduced the Deep Speech system when he was the chief scientist of Baidu. He said that in the past, different language-specific systems need to be implemented to identify different languages, and Deep Speech systems that use deep learning technology are completely unnecessary. Treat different languages ​​differently. "The versatility of learning algorithms has made it possible to distinguish between different languages."
Currently, Deep Speech has launched the second generation. Adam Coates said that the system was originally trained using English speech data and later introduced Chinese speech data training.
"Because this system is based entirely on deep learning, it depends entirely on training data, so we can quickly replace it with Chinese speech data and make it a very strong Chinese speech recognition engine," Coates said.
He explained that the training data of this system is very large, and it has added a lot of Chinese dialect data with strong accents, and the training effect of these data is also very good, so that many people have difficulty hearing the dialect voice system. Can be successfully identified.
Today, Baidu’s system, which was developed primarily by the US Institute of Artificial Intelligence, has been successfully productized. In October last year, a mobile application called TalkType landed on the Android Store. This application is based on Deep Speech technology and is Baidu’s first real “voice priority†input method instead of the keyboard. Input selection of "speech assist" other than input.
Coates believes that the biggest advantage of Baidu and other companies engaged in research on artificial intelligence in China is that Chinese Internet users quickly accepted such methods as Internet access for mobile terminals, whether it be ordinary e-mail, web browsing, or even dining in restaurants. All can be done with mobile devices. These usage habits are completely different from those of U.S. users, and these give a wide range of applications for artificial intelligence applications such as speech recognition.
3D printer stepper motors are important parts of 3D printers. They are used in a variety of applications depending on the type of printer. For example, 3D printer stepper motors are used to move the extruder or the build platform alone the x, y, and z axis. The extruder assembly also has a Stepper Motor used to pull the filament into the extruder.
3d printer stepper motor,Best stepper motor for 3d printer,Extruder stepper motor,3d printer extruder motor,Best motors for 3d printer,Stepper motor in 3d printer
Shenzhen Maintex Intelligent Control Co., Ltd. , https://www.maintexmotor.com