On November 22nd, Baidu announced the opening of four new voice technology interfaces to the public on the third anniversary of its voice open platform, so that users can enjoy a more wonderful interactive experience when using voice and machine communication.
Wu Ende, the chief scientist of Baidu, briefly introduced the four revolutionary voice technologies, namely emotional synthesis, far-field solution, wake-up phase II technology and long-voice solution, and announced that Baidu will open these technologies to users free of charge. Share with developers.
"These technologies have great potential to completely change the efficiency and methods of human-computer interaction. Future voice technology has a good opportunity in many application scenarios, which will bring huge changes to human-computer interaction." Wu Enda said.
These technologies are designed to address some of the key issues that users are generally troubled with when using voice interactions. For example, Baidu's emotional synthesis technology focuses on "adding emotions" to synthetic speech, which can now achieve close to real human voice effects. Baidu used this technology earlier this year to restore the voice of the deceased star Leslie Cheung.
Similarly, developers can use the new interface to increase the speech recognition distance to 3-5 meters, increase the device's voice wake-up rate to over 95% while saving less power and false positives, or improve long-term speech recognition. Accuracy issue. This will bring more imagination to voice technology than it is now, not just remote control or unlocking the phone.
For example, the representatives of the former two are Baidu's “small robot human-machine voice interactive ordering mealâ€, which has been put into use in the Shanghai KFC flagship store, and can answer the order at any time. The latter has already shown great imagination in many application scenarios such as content recording, intelligent customer service, and video transfer.
In this celebration called "The Sense of Openness and the Future of Common Languages", James Landay, an artificial intelligence expert from Stanford University, also shared a recent research collaboration with Baidu, which found that smartphones are input during speech. The input speed is 3 times faster than the keyboard input. He said, "In the past two years, thanks to the continuous development of big data and deep learning technology, speech recognition technology has advanced by leaps and bounds, and speed and accuracy have made great progress."
Wu Guilin opened the mobile application client of the video application "Iqiyi" and said "VIP renewal fee", the system accurately jumped to the corresponding recharge page. The iQiyi technical director pointed out that with the Baidu voice open platform, more than one million iQiyi users use voice search every day, and more than 80% of them are converted into effective clicks.
Reader Jin Dashi, general manager of Gansu Digital Technology Co., Ltd. believes that the value of the voice open platform is not limited to commercial. The “Reader Digital Farm Bookstore†has been successfully piloted in Qingyang City, Gansu Province, and completed 65 new rural “digital farmer's bookstoresâ€. He said, “Many illiterate elderly and left-behind children, speech synthesis allows them to enjoy reading too.â€
At present, the partners of Baidu Voice Open Platform have covered many fields and scenarios, including Lenovo and ZTE in the field of smart phones; Changhong Smart TV in smart home, Konka Smart TV, SONY Smart TV; Tesla and Tucson in the automotive industry. Hewlett-Packard, Amy Communications in the field of smart devices, Ctrip in smart services, and QQ reading on mobile phones.
"Voice is the most natural way of human communication. Through open voice technology, Baidu hopes to lead the prosperity of voice-enabled products." Wu Enda said.
It is reported that since the launch of the Baidu voice open platform in October 2013, the daily online voice recognition requirements have increased from 5 million in 2013 to 140 million today, and the daily request for online voice synthesis has reached 200 million. The number of developers exceeds 140,000.
In terms of technical indicators, Baidu's speech recognition accuracy rate has reached 97%, ranking first in the world. In February of this year, Baidu's deep speech recognition system Deep Speech 2 was selected as the top ten breakthrough technology of MIT 2016. Baidu Brain, including voice technology, was selected as one of the 15 leading scientific and technological achievements of the 3rd Wuzhen World Internet Conference in 2016, becoming the only comprehensive artificial intelligence technology selected in China.
Pressure Thermometer,Capillary Thermometer Measured,Capillary Thermometer Measured In Boiler,Stainless Steel Pressure Gauge
ZHOUSHAN JIAERLING METER CO.,LTD , https://www.zsjrlmeter.com