Video Summarization Using a Dense Captioning (DenseCap) Model
Sourav Das, Anup Kumar Kolya, Arindam Kundu · 2021
A huge number of video camera recordings produce a huge dataset. Dealing with this video data in terms of manpower, hardware storage, time, etc. is a complex task; therefore, a video-summarization technique is a much-needed algorithm. Video can be summarized in three different ways using keyframes, video skimming, or short segments. This chapter describes keyframe-based method and how it is executed. It presents a literature survey of the subject, discusses our approach, and details the methodology of implementation. The chapter discusses the implementation and the internal working details of DenseCap model. It reports the results of the experiments and compares the technique with other methods. Finally, the chapter discusses the limitations of the methodology and presents some conclusions.