TITAN：2024年生成式人工智能与虚假信息：最新进展、挑战和机遇（英文版）

Similar to text and visual synthesis, AI-based speech synthesis methods have made enormous progress in recent years, which is mainly due to the so-called “end-to-end learning” paradigm: Here, text analysis, acoustic modelling and audio synthesis are no longer isolated, but mapped, trained and optimised within a common network architecture. This network architecture can also be trained with roughly annotated data (in the form of pairs of text and audio recordings), which are available in almos