Unsupervised Deep Learning-based clustering for Human Activity Recognition

TL;DR

This paper introduces DISC, a deep learning-based clustering method that automatically labels unlabelled inertial sensor data for human activity recognition, addressing the challenge of limited labeled datasets.

Contribution

The paper presents a novel DL architecture combining a recurrent AutoEncoder with clustering to improve unsupervised HAR data labeling.

Findings

DISC outperforms four existing deep clustering methods in accuracy

Effective on three public HAR datasets

Improves normalized mutual information metrics

Abstract

One of the main problems in applying deep learning techniques to recognize activities of daily living (ADLs) based on inertial sensors is the lack of appropriately large labelled datasets to train deep learning-based models. A large amount of data would be available due to the wide spread of mobile devices equipped with inertial sensors that can collect data to recognize human activities. Unfortunately, this data is not labelled. The paper proposes DISC (Deep Inertial Sensory Clustering), a DL-based clustering architecture that automatically labels multi-dimensional inertial signals. In particular, the architecture combines a recurrent AutoEncoder and a clustering criterion to predict unlabelled human activities-related signals. The proposed architecture is evaluated on three publicly available HAR datasets and compared with four well-known end-to-end deep clustering approaches. The experiments demonstrate the effectiveness of DISC on both clustering accuracy and normalized mutual information metrics.