Detecting Unseen Falls from Wearable Devices using Channel-wise Ensemble of Autoencoders
This work addresses fall detection for elderly or at-risk individuals using wearable devices, presenting an incremental improvement in feature extraction and thresholding methods.
The paper tackled the problem of detecting unseen falls from wearable sensor data by proposing an ensemble of autoencoders trained only on normal activities, achieving improved detection performance over traditional methods on two activity recognition datasets.
A fall is an abnormal activity that occurs rarely, so it is hard to collect real data for falls. It is, therefore, difficult to use supervised learning methods to automatically detect falls. Another challenge in using machine learning methods to automatically detect falls is the choice of engineered features. In this paper, we propose to use an ensemble of autoencoders to extract features from different channels of wearable sensor data trained only on normal activities. We show that the traditional approach of choosing a threshold as the maximum of the reconstruction error on the training normal data is not the right way to identify unseen falls. We propose two methods for automatic tightening of reconstruction error from only the normal activities for better identification of unseen falls. We present our results on two activity recognition datasets and show the efficacy of our proposed method against traditional autoencoder models and two standard one-class classification methods.