Reproduce “Robust and lightweight audio fingerprint for Automatic Content Recognition” 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
Anoubhav Agarwaal, Prabhat Kanaujia, Sartaki Sinha Roy, Susmita Ghose
The authors propose a compact audio fingerprint to identify television content from short clips, using overlapping windows to make neighboring fingerprints change gradually and support a sparser reference index. In exhaustive search on Dataset-1, at skip 5 and 5 million stored fingerprints, it identifies 89.96% of clips versus 85.46% for min-hash; this test does not show temporal correlation alone causes the gain.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.