MUMBAI, India, Jan. 23 -- Intellectual Property India has published a patent application (202541117133 A) filed by Nandha Engineering College, Erode, Tamil Nadu, on Nov. 26, 2025, for 'real-time lecture transcription and ai-enhanced note generation.'
Inventor(s) include S Gobika; R L Jai Krishnaa; S Mahendhiran; and D Vinoparkavi.
The application for the patent was published on Jan. 23, under issue no. 04/2026.
According to the abstract released by the Intellectual Property India: "In today's digital learning environment, students often struggle to capture and understand crucial information from live lectures due to the fast pace of teaching, language barriers, and lack of structured study material. To address this challenge, we propose Class Scribe, a hybrid Al-powered system that performs realtime speech-to-text conversion, enhanced transcription, and intelligent note generation from live classroom audio and video content. The proposed system is built on a three-layer hybrid architecture consisting of a React Native mobile application, a Node js API Gateway, and a Python-based AI Processing Layer. During a lecture, the mobile application captures live audio and, in the extended version, can 11lso capture video frames to identify visual cues such as slides and written content on the board. The captured data is streamed to the backend using secure WebSocket communication. For real-time captions, the system uses the Vosk speech recognition engine, ensuring low-latency transcription even on low-resource environments. After the lecture ends, the recorded audio Is further processed using FasterWhisper to generate a more accurate full transcript with timestamps. This transcript is refined using punctuation restoration and grammar correction techniques. Important topics and key terms are extracted using Key BERT and spaCy, ensuring that the most relevant concepts are identified automatically. To enhance learning, Mistral-78, an open-source large language model running locally via Ollama, is used to generate detailed study notes from the refined transcript. These notes include topic-wise explanations, simple definitions, step-by-step breakdowns, real-world examples, and summarized key takeaways. This makes the output highly suitable for students preparing for exams or revision. As a unique feature, the system also integrates optional video context analysis, where important visual moments in the lecture (such as slide changes and handwritten diagrams) can be tagged and associated with the transcript. This enables students to revisit both the spoken and visual content together, creating a richer and more immersive learning experience. The entire system follows a microservices-based architecture, allowing independent scaling and efficient resource management. By combining real-time processing with deep AI understanding, ClassScribe bridges the gap between passive listening and active learning, offering a powerful tool for students, educators, and institutions aiming to improve classroom comprehension and study efficiency."
Disclaimer: Curated by HT Syndication.
