MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202617092743 A) filed by Google LLC on July 30, 2026, for Speech-Text Prompting For Speech Tasks.
Inventors include Kastner, Kyle Scott; Wang, Quan; Wang, Gary; Elias, Isaac; Rosenberg, Andrew; Ramabhadran, Bhuvana; Mengibar, Pedro J. Moreno; Park, Alex; Partridge, Kurt Edward; Haghani, Parisa; Agrawal, Rohan; Neeraj, Gaur; Yoon, Jae Yeun; Chun, Byungha; and Biadsy, Fadi.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: A method (500) includes receiving a reference utterance (302) and an input text utterance (304). The reference utterance includes a plurality of terms spoken by a reference speaker and the input text sequence includes a corresponding transcript (306) for each of the plurality of terms spoken by the reference speaker. The method includes obtaining a speaker embedding (308) characterizing speaker characteristics of the reference speaker. The method includes generating a replacement input text sequence (314) by replacing the corresponding transcript of a respective one of the plurality of terms with a replacement transcript (316) corresponding to a different term not included in the reference utterance. The method includes generating, using a text-to-speech (TTS) model (130) conditioned on the reference utterance and the speaker embedding, resynthesized speech (134) based on the replacement input text sequence in a voice of the reference speaker.
Disclaimer: Curated by HT Syndication.
