Reproduce “A Deep Siamese Network for Scene Detection in Broadcast Videos” 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
Lorenzo Baraldi, Costantino Grana, Rita Cucchiara
The authors develop a method that learns similarities between shots using visual content, transcript features, and timing. It converts distances into similarity scores for spectral clustering, which groups shots into scenes. On 11 Planet Earth episodes, it averages 0.48 on M_iou and 0.62 on F_co, versus 0.46 and 0.52 for color histograms with spectral clustering; gains are not consistent across episodes.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.