The team behind the project is also working on an "automatic speaker and emotion inference system" to make the dialogue more engaging. They hope to eventually allow users to select their own preferences for voice and speed, creating a custom audiobook. The lead researcher of the project has expressed hopes of creating free audiobooks for all 60,000 ebooks available on Project Gutenberg, possibly even translating them into different languages. The software used for this project is open source and available for others to use or enhance.
Key takeaways:
- Researchers from Microsoft, Google, and MIT have collaborated with Project Gutenberg to create 5,000 open-license audiobooks using a synthesized AI voice.
- The audiobooks are a mix of fiction and non-fiction, including classic literature, plays, and biographies, and are available on major podcast and streaming platforms.
- The AI voice is surprisingly human-like, although it does have some limitations, such as confusion with stand-alone letters and a lack of variety in voices.
- The team hopes to eventually create free audiobooks for all 60,000 of the ebooks available on Project Gutenberg, potentially even translating them into different languages.