MUMBAI, India, July 30 -- Intellectual Property India has published a patent application (202641088095 A) filed by Vellore Institute Of Technology on July 19, 2026, for A Portable Embedded System And Method For Real-Time Audiometric Classification Using Edge Intelligence.
Inventors include Sivakumar Rajagopal; and T Priyadharsini.
The application for the patent was published on July 24, 2026, under issue no. 30/2026.
Abstract: The present invention discloses a TinyML-enabled embedded audiometric classification apparatus and method for real-time hearing loss assessment using a low-power embedded microcontroller. The apparatus acquires frequency-based hearing threshold inputs corresponding to standard audiometric frequencies and processes the inputs through a feature processing module to generate a feature vector. A Random Forest classifier deployed as a TinyML inference engine performs on-device machine learning inference to classify hearing loss into predefined categories comprising Normal, Mild, Moderate, Severe, and Profound. The classification result is mapped to visual LED indicators and transmitted to a graphical user interface through a serial communication interface for real-time visualization. The apparatus further supports GUI-based dataset input and push-button-based audiometric simulation for interactive operation. By executing machine learning inference entirely on the embedded platform, the invention reduces latency, minimizes power consumption, eliminates dependence on cloud-based computation, and provides a portable, low- cost, and offline hearing assessment solution suitable for resource-constrained environments.
Disclaimer: Curated by HT Syndication.
