IntelliMeet: AI-powered meeting summarization with FLAN-T5 and cosine clustering
In response to increased use of virtual and hybrid meetings, IntelliMeet offers an AI-based solution for real-time transcription and summarization. We use Google’s Speech-to-Text API for high-quality transcription, abstractive summarization using fine-tuned FLAN-T5, and cosine clustering for information organization based on topics. With hybrid strategy using extractive and abstractive techniques, IntelliMeet provides readable and coherent meeting minutes. Experimental evaluations using SAMSum datasets prove better ROUGE-1 and ROUGE-2 ratings, displaying potential for large-scale use in corporate and academic environments.
Authors
- Ashish Awate (ORCID: https://orcid.org/0000-0002-5845-2363)
- Maneela Jain
- Bhushan Nandwalkar (ORCID: https://orcid.org/0000-0002-3387-6469)
- Soham Chaudhari
- Vaishnavi Pardeshi
- Pranali Magar
Institutions
- Shri Vile Parle Kelavani Mandal (IN)
Publication Details
- Journal
- IET conference proceedings.
- Published
- 2025-06-27
- DOI
- https://doi.org/10.1049/icp.2025.1566
- Primary Topic
- Semantic Web and Ontologies
- Type
- article
- Field-Weighted Citation Impact
- 0.00