MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202611087418 A) filed by Krishna Institute Of Engineering & Technology Kiet on July 16, 2026, for "system And Method For Automated Multilingual Audio-Visual Dubbing And Cultural Localization Using Adaptive Multi-Stage Neural Processing Pipelines".
Inventors include Bhagvan Krishna Gupta; Sparsh Singhal; Sudiksha Chauhan; Umang Saluja; and Samridh Raj.
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 automated multilingual audio-visual dubbing and cultural localization using adaptive multi-stage neural processing pipelines. The invention comprises video ingestion, audio extraction, human detection, speaker diarization, automatic speech recognition, translation, cultural adaptation, neural voice synthesis, lip synchronization, rendering, quality evaluation, and provenance verification modules. The system converts source- language audiovisual content into one or more target languages while preserving speaker identity, emotional expression, contextual meaning, and temporal synchronization. Adaptive synchronization architectures dynamically optimize facial movements according to synthesized speech. The invention further incorporates digital watermarking, authenticity verification, confidence-based quality routing, and feedback-driven learning mechanisms.
Disclaimer: Curated by HT Syndication.
