NLP-Driven Video Transcription: A Comprehensive Survey and Innovative Office Meeting Automation
P S Akshatha, B Akash, V Harshith, C Sumukh, V Gowardhan Reddy, Shravya Shetty · 2024
This paper introduces an innovative approach for automating office meeting summarisation, leveraging OpenAI’s Whisper ASR model and ChatGPT capabilities to enhance language processing. The Whisper ASR model adeptly transcribes spoken words from meetings into written text, effectively capturing the primary ideas exchanged during discussions. Subsequently, ChatGPT utilizes this transcribed text to generate concise and coherent summaries, emphasizing key points from the conversations. The system adeptly produces precise and meaningful meeting summaries through the synergistic integration of advanced speech recognition and language modelling, facilitating the extraction of crucial insights from spoken content. The system is thoughtfully crafted as a user-friendly web application, employing straightforward technologies such as HTML, CSS, and Flask for accessibility and ease of use. The intuitive interface enables participants to effortlessly upload video content from meetings, allowing the system to transform spoken words into easily comprehensible summaries in real time. Beyond the automation of summarization tasks, this proposed approach signifies a significant advancement in applying language technologies to enhance communication efficiency in professional settings. By seamlessly combining cutting-edge video transcription, text summarization, and web development, the project not only streamlines the summarization process but also paves the way for more dynamic and effective information sharing within office environments. As this work is part of a survey paper, the proposed idea contributes to the broader exploration of leveraging language technologies for improved communication and information exchange.