m monad-knowledge FIIT STU · open research vault
dataset

aril

Also known as: ARIL, ARIL dataset, Activity Recognition with Indoor Localization

Description

ARIL (Activity Recognition with Indoor Localization) is a public WiFi CSI dataset jointly labelled with activity class and indoor location. It is used as a benchmark for multi-task CSI models that predict an activity and a coarse spatial label from the same trace, and shows up across CSI HAR and self-supervised learning papers as a cross-task evaluation set.

Modality / size

  • Modality: WiFi CSI from commodity NICs (Intel 5300 lineage).
  • Subjects / scenarios: indoor lab setting with multiple location bins.
  • Labels: paired activity label + location label per sample.
  • Exact sample count: see source paper.

Used by (papers)

  • Standard joint activity + localisation benchmark in the CSI sensing literature.
  • Frequently paired with UT-HAR, NTU-Fi, and Widar3.0 in comparative CSI HAR studies.

Notes

  • Public dataset, released alongside the ARIL paper.
Where it appears

3 vault papers evaluate on this dataset

Titles and DOIs only — no abstracts, no analyses.

  • WiMANS: A Benchmark Dataset for WiFi-based Multi-user Activity Sensing 2025 DOI ↗
  • A Survey on Green Wireless Sensing: Energy-Efficient Sensing via WiFi CSI and Lightweight Learning 2026 DOI ↗
  • A survey on CSI-based Wi-Fi sensing datasets and models with a focus on reproducibility 2026 DOI ↗
Co-mentioned with
Convolutional Neural Network×3 Activity Recognition×3 CSI Fingerprinting×3 Indoor Localization×3 WiFi CSI Sensing×3 Neural Network×3 domain generalization×2 Gesture Recognition×2 Privacy-Preserving Sensing×2 Environment Dependence×2