On Multi-resident Activity Recognition in Ambient Smart-Homes

TL;DR

This paper evaluates various methods for multi-resident activity recognition in smart homes, highlighting the effectiveness of recurrent neural networks with gated units and combined activity labels for improved accuracy and efficiency.

Contribution

It provides a comprehensive benchmark and comparison of different models for multi-resident activity recognition, emphasizing the superiority of RNNs with gated units and combined labels.

Findings

RNNs with gated recurrent units outperform other models.

Using combined activity labels is more effective than separate labels.

Gated recurrent units are both accurate and computationally efficient.

Abstract

Increasing attention to the research on activity monitoring in smart homes has motivated the employment of ambient intelligence to reduce the deployment cost and solve the privacy issue. Several approaches have been proposed for multi-resident activity recognition, however, there still lacks a comprehensive benchmark for future research and practical selection of models. In this paper we study different methods for multi-resident activity recognition and evaluate them on same sets of data. The experimental results show that recurrent neural network with gated recurrent units is better than other models and also considerably efficient, and that using combined activities as single labels is more effective than represent them as separate labels.