README.md

metadata

title: LSPW Periodic Workflow Discovery
emoji: 🔄
colorFrom: blue
colorTo: indigo
sdk: gradio
sdk_version: 6.2.0
app_file: app.py
pinned: false

Unsupervised Discovery of Long-Term Spatiotemporal Periodic Workflows in Human Activities

Project Page | arXiv

Abstract

Periodic human activities with implicit workflows are common in manufacturing, sports, and daily life. While short-term periodic activities—characterized by simple structures and high-contrast patterns—have been widely studied, long-term periodic workflows with low-contrast patterns remain largely underexplored.

To bridge this gap, we introduce the first benchmark comprising 580 multimodal human activity sequences featuring long-term periodic workflows. The benchmark supports three evaluation tasks aligned with real-world applications: unsupervised periodic workflow detection, task completion tracking, and procedural anomaly detection. We also propose a lightweight, training-free baseline for modeling diverse periodic workflow patterns.

Usage

Dependencies

Ensure you have the following Python packages installed:

• numpy
• scikit-learn
• tqdm
• matplotlib
• scipy

You can install them using pip:

pip install numpy scikit-learn tqdm matplotlib scipy

Estimation

Run the workflow detection function to perform unsupervised periodic workflow detection on the dataset.