Reproduce “Asking a Multimodal LLM Whether Someone Fell: Zero-Shot Screening of Low-Resolution Thermal Video in a Laboratory and an Occupied Home” 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
Marcos Junca, Mauro Junca
The system uses event-triggered thermal video and a multimodal model to review short clips for falls without task-specific training. Gemini 3.8 Flash called 69% of held-out laboratory falls and none of 499 non-fall windows; across 76 hours in one home, it made no false calls but identified only 4 of 9 staged, camera-visible falls, all staged by younger adults.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.