Shulou(Shulou.com)11/24 Report--

CTOnews.com, Sept. 12 (Xinhua) Microsoft recently partnered with the "free digital library project" Project Gutenberg to produce more than 5000 free audio books for the e-library using AI.

According to the ▲ image source Project Gutenberg website, in this project, researchers have combined a number of AI technologies, such as machine learning, automatic text selection and natural speech synthesis, to realize the whole process of automatically converting e-books into speech.

After querying, CTOnews.com learned that the project team first developed an algorithm that can automatically analyze the structure of e-books, distinguish between page numbers and tables of text and non-critical parts, and then use text-to-speech technologies such as WaveNet, Tacotron and FastSpeech to generate natural speech similar to human beings. In addition, the team also enables the system to identify the narrator, dialogue and different characters in the book, and adjust the voice accordingly to achieve the voice expression of the storyline.

Microsoft said that the program has collected more than 35000 hours of audio book voice data, covering a variety of classic literature, dramas, biographies and other works, which will be provided free of charge as an open source. Users can also use a small amount of voice recorded by themselves to create the effect of "reading the whole book with their own voice". This project relatively enriches the content of audio books on the website and provides convenience for the visually impaired.

Referenc

Thousands of free and open audiobooks powered by Microsoft AI

Project Gutenberg official website

Paper Large Scale Intelligent Microservices

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.