Newest papers

Asking a Multimodal LLM Whether Someone Fell: Zero-Shot Screening of Low-Resolution Thermal Video in a Laboratory and an Occupied Home

Marcos Junca, Mauro Junca · 2026 · arXiv

DOI: 10.48550/arXiv.2610.29913 · open-access PDF

Portuguese

As câmeras térmicas podem monitorar as pessoas mais velhas que vivem sozinhas sem capturar rostos, e os modelos multimodais de grande linguagem (LLMs) podem julgar um vídeo curto sem treinamento específico de tarefas. Avaliamos um LLM multimodais de produção, Gemini 3.8 Flash a temperatura 0, como a fase de decisão de um screener de queda desencadeado por eventos, usado exatamente como um protótipo implantado: um pedido por clipe térmico curto, com sua trilha sonora quando há um, e seu antecessor Gemini 3.7 Flash sobre os mesmos pedidos. Em 438 quedas de estágio e 563 janelas de não queda que colocamos à mão em um conjunto de dados térmicos público 32x24

Machine translation (Meta M2M100); the English original follows.

Abstract

Thermal cameras can watch over older people living alone without capturing faces, and multimodal large language models (LLMs) can judge a short video without task-specific training. We evaluate a production multimodal LLM, Gemini 3.8 Flash at temperature 0, as the decision stage of an event-triggered fall screener, used exactly as a deployed prototype uses it: one request per short thermal clip, with its sound track when there is one, and its predecessor Gemini 3.7 Flash on the same requests. On 438 staged falls and 563 non-fall windows that we located by hand in a public 32x24 thermal dataset

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Is this paper real? This page was built from OpenAlex and arXiv records, not generated. Software can verify any reference in one call: GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.29913, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txt

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange