MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202641049595 A) filed by Seshadri Rao Gudlavalleru Engineering College; Potturi Reshma; Mohammad Ali Akmal Baig; Mallavalli Sai Vivek; Shaik Jameer Ahmadh; and Parisa Deepthi, Gudlavalleru, Andhra Pradesh, on April 18, for 'automatic generation of segmented educational videos from text using nlp and generative ai.'
Inventor(s) include Seshadri Rao Gudlavalleru Engineering College; Potturi Reshma; Mohammad Ali Akmal Baig; Mallavalli Sai Vivek; Shaik Jameer Ahmadh; and Parisa Deepthi.
The application for the patent was published on May 1, under issue no. 18/2026.
According to the abstract released by the Intellectual Property India: "The increasing demand for engaging and accessible digital learning content has highlighted the limitations of traditional text-based educational materials. This paper presents an AI-based system that automatically converts educational text documents into segmented visual narration videos using a fully automated multimodal pipeline. The suggested structure combines document parsing, Natural Language Processing (NLP), Large Language Models (LLMs), Text- to-Speech (TTS), and diffusion-based Text-to Image (TTI) generation to convert flat documents into organized narrated video segments. Educational material is handled at the sentence granularity to facilitate detailed alignment of narration, images and captions. A robust primary-fallback strategy is employed across LLM, TTS, and image generation modules to ensure reliable operation. Experimental evaluation demonstrates that the system produces coherent and pedagogically suitable educational videos while significantly reducing manual effort and content production time through full pipeline automation. The proposed solution is scalable, cost-efficient, and well suited for e-learning platforms, educators, and inclusive digital education delivery platforms."
Disclaimer: Curated by HT Syndication.
