What You Say Is Not What You Do: Studying Visio-Linguistic Models for TV Series Summarization | IEEE Conference Publication | IEEE Xplore