WHAR Datasets: An Open Source Library for Wearable Human Activity Recognition

2025-08-12

AI Summary

WHAR research has long suffered from fragmented dataset formats, leading to poor reproducibility, low cross-study comparability, and inefficient preprocessing. To address this, we introduce the first standardized, open-source data processing library for WHAR, featuring a unified data interface and configuration-driven architecture compatible with PyTorch and TensorFlow, and integrating nine mainstream WHAR datasets. Our method employs declarative configuration to fully automate data loading, segmentation, augmentation, and batching, and introduces a multi-process parallel preprocessing engine. Experiments demonstrate up to 3.8× faster preprocessing versus manual implementations. We successfully reproduce TinyHar and MLP-HAR with performance matching original reports, validating the library’s advantages in reproducibility, extensibility, and computational efficiency. This work establishes a standardized infrastructure to advance WHAR research.

Technology Category

• Machine Learning: Hardware-aware ML
• Humans and AI: Human-in-the-loop Machine Learning
• Computer Vision: Segmentation

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Data management and stream processing for Web, mobile and wireless applications
• Web Mining and Content Analysis: Large pretrained models with web data
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

The lack of standardization across Wearable Human Activity Recognition (WHAR) datasets limits reproducibility, comparability, and research efficiency. We introduce WHAR datasets, an open-source library designed to simplify WHAR data handling through a standardized data format and a configuration-driven design, enabling reproducible and computationally efficient workflows with minimal manual intervention. The library currently supports 9 widely-used datasets, integrates with PyTorch and TensorFlow, and is easily extensible to new datasets. To demonstrate its utility, we trained two state-of-the-art models, TinyHar and MLP-HAR, on the included datasets, approximately reproducing published results and validating the library's effectiveness for experimentation and benchmarking. Additionally, we evaluated preprocessing performance and observed speedups of up to 3.8x using multiprocessing. We hope this library contributes to more efficient, reproducible, and comparable WHAR research.

Problem

• Lack of standardized formats for wearable activity datasets
• Need for reproducible and efficient data handling workflows
• Difficulty in comparing results across different HAR datasets

Innovation

• Standardized data format for wearable activity recognition
• Configuration-driven design enabling reproducible workflows
• Integration with PyTorch and TensorFlow frameworks