Hierarchical and Multimodal Data for Daily Activity Understanding

\name Ghazal Kaviani \email [email protected], \name Yavuz Yarici \email [email protected], \name Seulgi Kim \email [email protected], \name Mohit Prabhushankar \email [email protected], \name Ghassan AlRegib \email [email protected], \name Mashhour Solh \email [email protected], \nam · 2025

Read Hierarchical and Multimodal Data for Daily Activity Understanding by \name Ghazal Kaviani \email [email protected], \name Yavuz Yarici \email [email protected], \name Seulgi Kim \email [email protected], \name Mohit Prabhushankar \email [email protected], \name Ghassan AlRegib \email [email protected], \name Mashhour Solh \email [email protected], \nam (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-17696.txt.

Introduction

Kaviani and six co-authors released a dataset for teaching machines to understand daily human activities in twenty twenty-five. Recognizing what people do in real homes is hard because activities nest inside each other and sensors disagree. Their dataset, DARai, records fifty participants in ten environments, scripted and unscripted.

Over two hundred hours came from twenty sensors: multiple cameras, depth and radar units, wearable motion trackers, muscle sensors, pressure insoles, biomonitors, and gaze trackers. Annotations form three levels, from whole activities down through shared actions to fine execution steps. Deliberate overlap between levels creates counterfactual combinations for stress-testing models. Experiments covered recognition, timing localization, and anticipating future actions, with single-sensor and fused setups. The authors present DARai as a benchmark exposing current models' limits.

Catalog record

CREATOR\name Ghazal Kaviani \email [email protected], \name Yavuz Yarici \email [email protected], \name Seulgi Kim \email [email protected], \name Mohit Prabhushankar \email [email protected], \name Ghassan AlRegib \email [email protected], \name Mashhour Solh \email [email protected], \nam
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE120 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-17696

Browse the full library · About the Tower of Babel