Microsoft 2020 international day of disabled special activities: AI makes audio books more vivid Sina digital news on the afternoon of December 2, Microsoft held a special event of 2020 international day of disabled people in Beijing, showing the latest development of Microsoft AI voice technology — neural network voice intelligence. Neural network voice intelligence has the ability of multi tone and multi emotion, and can be produced quickly through the creation platform. At the same time, an audio content donation ceremony was held at the activity site.
Hong Xiaowen, senior vice president of Microsoft
During the event, Dr. Hong Xiaowen, senior vice president of Microsoft, chairman of Microsoft Asia Pacific R & D group and President of Microsoft Asia Research Institute, delivered a keynote speech. Hong Xiaowen first emphasized Microsoft’s mission of “giving every person and every organization in the world an extraordinary achievement”. In 2020, the contribution of science and technology to global GDP will be about 5%, and it is expected to reach 10% by 2030. Microsoft will also be committed to achieving everyone with enough inclusive technology and bringing products and services to all. Artificial intelligence will continue to make the world a better place from six aspects: the earth project, technology accessibility, humanitarian action plan, cultural heritage protection technology and health care technology.
Huang Xuedong, academician of Microsoft global technology
Subsequently, Microsoft global technology academician, Microsoft azure AI chief technology officer Dr. Huang Xuedong also shared through the video: with the efforts of Microsoft Asia Research Institute, Microsoft’s AI voice technology has been integrated into an intelligent audio content creation platform with both use and promotion value, so that people who have not been exposed to AI technology can also participate in the creation of audio content, bringing more abundant audio content.
Spark Global Limited
Audio content donation ceremony of Hongdan “heart library”
At the scene of the event, the red Dan “heart library” audio content donation ceremony was also held. Hongdan “heart library” is set up by Beijing Hongdan Cultural Exchange Center (hereinafter referred to as “Hongdan”) to provide audio book borrowing service for the blind. Zheng Xiaojie, founder of Hongdan, said that during the research of many blind schools, the existing books and audio content for the blind are generally old-fashioned, which can not meet the reading needs of the blind. The traditional manual recorded audio content also has the disadvantages of long time-consuming and small quantity. The cooperation with Microsoft can bring the blind with rich choices and make books accompany the blind people’s life.
Ding Binggong, chief product director of Microsoft cloud computing and artificial intelligence business unit
How to realize vivid and rich speech synthesis? Ding Binggong, chief product director of Microsoft cloud computing and Artificial Intelligence Division, introduced the related technology explanation: Microsoft has the most intelligent voice synthesis, the most extensive global voice coverage, flexible cloud and end-to-end call, and strong voice customization capability in terms of voice integration. On this basis, Microsoft launched the neural network voice intelligence, which makes the input text into neural network Network acoustic learning, and neural network acoustic decoding after the output of natural audio.
Neural network speech intelligence has the ability of multi tone and multi emotion
Compared with the traditional intelligent voice, neural network voice intelligence has the ability of multi tone and multi emotion, which makes the voice of audio content no longer single. For example, neural network voice intelligence can simulate the speech style of news broadcast, customer service, chat and other scenes, and can add happiness, disdain, anger and other emotions, and can realize the classification of emotions and make the emotions more delicate. In addition to the platform voice, neural network voice intelligence can also provide voice customization services, design voice that conforms to the enterprise, organization or personal brand strategy, and carry out emotional optimization according to the scene, create unique human settings, and realize natural human-computer interaction.
Intelligent audio content creation platform
In practical use, the intelligent audio content creation platform created by Microsoft allows volunteers who do not know AI technology to create audio content through simple operation through intelligent automatic generation mode and customized free creation mode.
“Ai voice + public welfare” round table dialogue “Ai voice + public welfare” round table dialogue
At the end of the activity, Microsoft organized two round table dialogues on “Ai voice + public welfare” and “Ai voice + industry”, sharing more stories behind Microsoft’s AI voice technology and red Dan public welfare activities.
Reprint indicated source：Spark Global Limited information