Reproduce “Intellectual Property Protection of Hyperdimensional Computing Classifiers: Black-Box Ownership Verification via Model Watermarking” 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
Hamed Poursiami, Farshad Soleiman, Ahmad Tahmasivand, Guojing Cong, Thomas Potok, Khaled N. Khasawneh, Kai Zeng, Maryam Parsa
The authors adapt backdoor watermarking to hyperdimensional computing classifiers by training secret out-of-distribution triggers with target labels, then verifying them through black-box predictions. At 1% exposure, AdaptHD and OnlineHD pass both verification thresholds on four benchmarks, while Vanilla HDC fails. On ISOLET, OnlineHD achieves 93.29% test accuracy and 48.3/50 trigger matches; tested extraction surrogates lose the trigger associations.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.