MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202641107182 A) filed by Easwari Engineering College; and Srm Institute Of Science And Technology,ramapuram Campus on September 07, 2026, for Semantic-Aware Personalized Voice Style Transfer Using Speech Diffusion Transformer.
Inventors include Kailas Dev R R; Balasubramaniyam N; and Ms. Preethi B.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: The present invention relates to a system and method for semantic-aware personalized voice style transfer using a speech diffusion transformer. The system receives reference speech from a target speaker and derives speaker identity and speaker-specific expressive behaviour, including fundamental frequency, pitch range, speaking rate, vocal energy, rhythmic timing, intonation, stress, duration, and pause patterns. Input text is analysed to determine semantic and emotional information required for expressive speech synthesis. The speaker-specific expressive behaviour is combined with semantic-emotional representation through behaviour-semantic fusion to generate a conditioning representation. The representation is utilized for personalized prosody prediction and conditioning of speech diffusion transformer during speech generation. Generated speech is processed through expressive control and neural decoding to produce a waveform preserving target speaker identity and expressive behaviour while reflecting semantic-emotional context. Total Number of Words in Abstract: 129
Disclaimer: Curated by HT Syndication.
