MUMBAI, India, Nov. 21 -- Intellectual Property India has published a patent application (202521103986 A) filed by Symbiosis Internationa, Pune, Maharashtra, on Oct. 28, for 'offline voice-activated smart glasses system for real-time visual scene captioning and auditory assistance.'
Inventor(s) include Sahil Goyal; Vaibhav Sharma; Girish Mahale; Jayaditya Shukla; Mayur Gaikwad; Shripad Deshpande; Nilkanth Deshpande; Shivali Wagle; Nandhini K; and Harikrishnan R.
The application for the patent was published on Nov. 21, under issue no. 47/2025.
According to the abstract released by the Intellectual Property India: "The present invention relates to a fully offline, voice-activated wearable smart glasses system designed to provide real-time visual scene captioning and natural-sounding speech output for visually impaired users. The system comprises an INMP441 MEMS microphone for wake-word detection, an ESP32-Wroom-32 microcontroller for initiating image capture through an ESP32-CAM with an OV2640 camera, and a Raspberry Pi 5 for executing an AI inference module that integrates an EfficientNetB4 encoder and Transformer decoder for caption generation. The generated captions are converted into audio using an offline Text-to-Speech engine, and output through an amplifier and speaker assembly. Operating entirely without internet connectivity, the system ensures low latency, privacy preservation, and energy efficiency. With real-time processing under 450 milliseconds, ergonomic design, and over six hours of continuous use, the invention enables autonomous navigation, contextual awareness, and enhanced accessibility for visually impaired individuals in diverse real-world environments."
Disclaimer: Curated by HT Syndication.
