 This paper proposes novel methods for extracting keyframes from digital video content stored in data centers. These keyframes can then be used to generate short summaries of the video content. The paper also evaluates the performance of several deep learning models on four popular video summarization data sets. The results show that the proposed solutions outperform existing methods, achieving higher F-scores than other methods. This article was authored by Obata ESA and Tamer Shanable.