Assessing Distribution Shift in Human Activity Recognition for Domain Generalization
Researchers have identified challenges in Human Activity Recognition (HAR) models due to data diversity from device and sensor heterogeneity, and introduced new benchmark platforms and datasets to address domain generalization.
HAR models struggle with data diversity from device and sensor heterogeneity, according to a study submitted in 2026[1]. Distribution shifts in HAR models can be caused by variations in device type, sensor placement, sampling rate, and user behavior. A new benchmark platform and datasets have been introduced to spur further research on domain generalization in HAR. Meanwhile, a separate study found that multi-modal deep learning models outperform uni-modal counterparts in HAR from wearable sensors[2]. The HARMES dataset, comprising 61 hours of fully labeled IMU, audio, and ambient humidity data, focuses on 15 household and personal hygiene activities of daily living (ADLs). Gated Multi-modal Fusion achieved the highest macro F1-score (0.82) in the comparison, surpassing the concatenation-based late fusion HARMES paper baseline by +6pp under leave-one-participant-out evaluation.
research-paperbenchmarktool-releasecommentary
Background sources we checked (4)
- arxiv.org ↗ While the field of Human Activity Recognition (HAR) continues to draw interest from researchers and advance in important ways, some key challenges remain. One of the most difficult aspects of building HAR models that show good performance in real-world settings is dealing with da…
- en.wikipedia.org ↗ Machine learning (ML) is a field of study in artificial intelligence concerned with the development and study of statistical algorithms that can learn from data and generalize to unseen data, and thus perform tasks without being explicitly programmed. Advances in the field of de…
- en.wikipedia.org ↗ A convolutional neural network (CNN) is a type of feedforward neural network that learns features via filter (or kernel) optimization. This type of deep learning network has been applied to process and make predictions from many different types of data including text, images and …
- en.wikipedia.org ↗ In machine learning, a neural network (NN) or neural net, is a computational model inspired by the structure and functions of biological neural networks. A neural network consists of connected units or nodes called artificial neurons, which loosely model the neurons in the brain.…