Meeting video analysis and summary generation system

Abstract

The present disclosure relates to a system that automatically extracts important information from a meeting video and generates a summary. A speech recognition unit extracts audio data from the video, and performs noise removal and speaker separation to convert it into text. A natural language processing unit performs morphological analysis, syntactic analysis, and semantic analysis on the text data to extract important keywords and phrases. A machine learning unit learns past meeting data and identifies important topics and key points of discussion. A summary generation unit creates a summary based on these key points and provides it in a format that a user can quickly understand. Furthermore, a video clip generation unit clips an important part of the video clip so that the user can directly watch it.

Dependencies are inferred from the claim text.