MUMBAI, India, Sept. 7 -- Intellectual Property India has published a patent application (202641104395 A) filed by Srm Trp Engineering College on August 31, 2026, for A Portable Edge-Based Offline Ai System For Code-Mixed Low-Resource Tribal Speech Recognition, Context-Aware Translation And English Voice Synthesis Using Mfcc-Cnn.
Inventors include Meena Dharrsini S; Lekha Shree B; Maran S; Rajarajeswari T; and Dr. P. Sudhakaran.
The application for the patent was published on September 04, 2026, under issue no. 36/2026.
Abstract: Tribal languages are often underrepresented in modern speech recognition technologies due to limited datasets and computational resources. This invention presents TribaLingua, an offline speech recognition and translation system designed for low-resource tribal languages. The system captures speech input from tribal language speakers and processes the audio using Mel Frequency Cepstral Coefficient (MFCC) feature extraction and a Convolutional Neural Network (CNN) based classification model. Recognized tribal phrases are mapped to English text and converted into audible speech using an offline Text-to-Speech (TTS) engine. The system operates entirely offline and can be deployed on low-power devices, making it suitable for use in rural and tribal regions with limited internet connectivity. The invention promotes digital inclusion, communication accessibility and preservation of indigenous languages.
Disclaimer: Curated by HT Syndication.
