MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202611110411 A) filed by Central University Of Punjab on September 15, 2026, for A Real-Time Unified Assistive Speech-To-Text And Text-To-Speech Technology For Specially Abled Students For Learning Environments.
Inventors include Dr. Meenakshi; Mohammed Ashif. C. K; Dr. Parneet Kaur; Dr. Shweta Sharma; and Dr. Kamaldeep.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: A unified assistive communication system (100) and method (200) for providing real-time accessibility support in educational environments are disclosed. The system (100) comprises a microphone (102), a camera (104), and a processor (106) configured to selectively operate in a Speech-to-Text (STT) processing mode and a Text-to- Speech (TTS) processing mode. In the STT mode, speech signals captured through the microphone are processed using a transformer-based speech recognition model comprising a Whisper-based encoder-decoder architecture adapted using a Low-Rank Adaptation (LoRA) mechanism. The processor further performs phonetic correction using Double Metaphone encoding and similarity-based correction using a domain-specific vocabulary database. In the TTS mode, images containing textual information captured through the camera are processed using optical character recognition, edit-distance-based candidate generation, and contextual correction using a bigram language model. Corrected text is converted into synthesized speech output. FIG. 3 will be the reference figure.
Disclaimer: Curated by HT Syndication.
