Guardado en:
Detalles Bibliográficos
Autores principales: Xu, Huatao, Tong, Panrong, Li, Mo, Srivastava, Mani
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:https://arxiv.org/abs/2412.15714
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912167133970432
author Xu, Huatao
Tong, Panrong
Li, Mo
Srivastava, Mani
author_facet Xu, Huatao
Tong, Panrong
Li, Mo
Srivastava, Mani
contents This paper introduces a novel mobile sensing application - life journaling - designed to generate semantic descriptions of users' daily lives. We present AutoLife, an automatic life journaling system based on commercial smartphones. AutoLife only inputs low-cost sensor data (without photos or audio) from smartphones and can automatically generate comprehensive life journals for users. To achieve this, we first derive time, motion, and location contexts from multimodal sensor data, and harness the zero-shot capabilities of Large Language Models (LLMs), enriched with commonsense knowledge about human lives, to interpret diverse contexts and generate life journals. To manage the task complexity and long sensing duration, a multilayer framework is proposed, which decomposes tasks and seamlessly integrates LLMs with other techniques for life journaling. This study establishes a real-life dataset as a benchmark and extensive experiment results demonstrate that AutoLife produces accurate and reliable life journals.
format Preprint
id arxiv_https___arxiv_org_abs_2412_15714
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle AutoLife: Automatic Life Journaling with Smartphones and LLMs
Xu, Huatao
Tong, Panrong
Li, Mo
Srivastava, Mani
Artificial Intelligence
Computation and Language
Human-Computer Interaction
This paper introduces a novel mobile sensing application - life journaling - designed to generate semantic descriptions of users' daily lives. We present AutoLife, an automatic life journaling system based on commercial smartphones. AutoLife only inputs low-cost sensor data (without photos or audio) from smartphones and can automatically generate comprehensive life journals for users. To achieve this, we first derive time, motion, and location contexts from multimodal sensor data, and harness the zero-shot capabilities of Large Language Models (LLMs), enriched with commonsense knowledge about human lives, to interpret diverse contexts and generate life journals. To manage the task complexity and long sensing duration, a multilayer framework is proposed, which decomposes tasks and seamlessly integrates LLMs with other techniques for life journaling. This study establishes a real-life dataset as a benchmark and extensive experiment results demonstrate that AutoLife produces accurate and reliable life journals.
title AutoLife: Automatic Life Journaling with Smartphones and LLMs
topic Artificial Intelligence
Computation and Language
Human-Computer Interaction
url https://arxiv.org/abs/2412.15714