Reproduce “Zygarde: Time-Sensitive On-Device Deep Inference and Adaptation on Intermittently-Powered Systems” 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

Bashima Islam, Shahriar Nirjon

Zygarde enables batteryless sensors to classify inputs under intermittent power by dividing neural-network inference into mandatory and optional work. K-means classifiers assess whether an input can exit early, while a scheduler uses deadlines and harvesting predictability to allocate computation. In a camera experiment, 93% of entered jobs met deadlines, but 37% of events never entered; mandatory completion is not guaranteed.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.