Using Fan-Made Content, Subtitles and Face Recognition for Character-Centric Video Summarization

Ismail Harrando, Alison Reboud, Pasquale Lisena, Raphaël Troncy, Jorma Laaksonen, Anja Virkkunen, Mikko Kurimo
2020 TREC Video Retrieval Evaluation  
This paper describes a fan-driven and character-centered approach proposed by the MeMAD team for the 2020 TRECVID [Awad et al. 2020] Video Summarization Task. Our approach relies on fan-made content and, more precisely, on the BBC EastEnders episode synopses from its Fandom Wiki 1 . This additional data source is used together with the provided videos, scripts and master shot boundaries. We also use BBC EastEnders characters' images crawled from the Google search engine in order to train a face
more » ... recognition system. All our runs use the same method, but with varying constraints regarding the number of shots and the maximum duration of the summary. The shots included in the summaries are the ones whose transcripts and visual content have the highest similarity with sentences from the synopsis. The runs submitted are as follows:
dblp:conf/trecvid/HarrandoRLTLVK20 fatcat:dihlrb7bxbesvnvuxjvldrlpz4