Reproduce “"Previously on ..." From Recaps to Story Summarization” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Aditya Kumar Singh, Dhruv Srivastava, Makarand Tapaswi
TaleSumm uses recap footage as weak supervision to identify shots and dialog in television episodes. It matches recap frames to episode shots, expands matches to nearby material, and scores shots and utterances using context. On 24’s IntraCVT test set, its model reaches 50.1 ± 2.8 video AP and 46.0 ± 2.1 dialog AP on two crime-thriller series.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.