10th International Congress on Information and Communication Technology in concurrent with ICT Excellence Awards (ICICT 2025) will be held at London, United Kingdom | February 18 - 21 2025.
Authors - Aryan Kasare, Mohd Kaif Idrisi, Asif Iqbal, Dipti Jadhav Abstract - Video Summarization (VS) is a methodology that highlights the main contents or events from the entire video in the form of keyframes. This paper presents a two-staged intelligent dynamic video summarization technique. In the first stage, SURF keypoints are extracted from the video frames and are given as input to a dictionary learning algorithm to identify the keyframes. In the second stage, the aim is to optimize the Stage-1 dynamic video summary using a multi-objective genetic algorithm to remove any redundancy in the generated video summary. The video summary results of the videos from the SUMME dataset demonstrate the validity of the proposed work. The subjective and objective performance analysis values justify the efficiency of the proposed algorithms to generate an optimized video summary.
