IIIT-Hyderabad develops AI tool to make science videos

IIIT-Hyderabad students and faculty have developed an AI-powered tool that converts complex scientific research papers into easy-to-understand video summaries in 11 languages. The project, in partnership with Anusandhan NRF, aims to make science accessible to all learners

Hyderabad: Students and faculty of the International Institute of Information Technology (IIIT) – Hyderabad have developed an Artificial Intelligence tool, which is set to change how science research is understood even by non-science students.

The IIIT-Hyderabad created a tool that not just simplifies the complex science research papers but also makes engaging and concise video presentations in any one of 11 languages, including English.

---

For this, the IIIT-Hyderabad team tied up with Anusandhan National Research Foundation, which promotes and fosters research and development, and innovation culture across the universities and colleges in the country.

Inspired by Google’s NotebookLM, which summarises uploaded documents such as research papers and converts them into easy-to-understand podcasts, the team comprising project leaders Dr. Lakshmanan Natraj and Rahul Sundar, along with developers Imandi Sai Ganesh, Arihant Rastogi and Vishnu Sathwik, went a step ahead and came up with video format.

According to the IIIT-Hyderabad team, a three-four minutes video presentation can be done in one of three ways – directly via LaTex files that contain the source codes and figures, by importing the paper from arXiV using the URL or by uploading the paper’s PDF.

The tool then breaks down content into individual slides that mirror the sections that an academic paper typically contains – Introduction, Methodology, Results, Discussion and a Conclusion – allowing the user to customise the bulleted points on each slide.

It also allows the user to edit the content. From the created slide deck, the last step enables the user to create a corresponding video and audio in any one of 11 languages, including English.

“The research summary is created with the help of AI tools such as Gemini, Claude, GPT and so on. When the script is ready in the form of slides, we then use Sarvam’s Text-to-Speech conversion engine to create an audio in the voice (either male or female) and language of our choice,” said Prof. Ponnurangan Kumaraguru, who led the efforts from IIIT-Hyderabad.