The first AI virtual anchor in Jiangsu Province was officially launched at a news app in Nanjing. Through intelligent synthesis technology, the AI virtual anchor could broadcast as a human being and what is more, she can speak 14 languages.
Realistic image, natural sound, precise mouth shape, no one can think of this as a virtual AI image. The AI virtual anchor is made possible by using face modeling technology to capture facial movements. The virtual anchor can control its own expressions according to the voice and match the appropriate movements to achieve more realistic effects.
Compared with real people, virtual anchors can be broadcast online in real time and at any time. They can also switch between multiple styles and multiple languages. They speak 14 languages, including Chinese, English, Japanese, Korean, Russian and French. These powerful functions are made possible by speech synthesis technology.
By collecting and recording the sound material of the real person, and then using the sound annotation and the machine's deep learning algorithm, we can construct her acoustic model of the pronunciation. On this basis, by inputting any text, she can make online broadcast to ensure the timely, fresh and disseminative news. Don’t you think it awesome?
(Source:ourjiangsu.com)
