Robust Human Activity Recognition based on Deep Metric Learning

Mubarak G. Abdu-Aguye, Walid Gomaa

Abstract: In the domain of Activity Recognition, the proliferation of low-cost and sensor-enabled personal devices has led to significant heterogeneity in the data generated by users. Traditional approaches to this problem have previously relied on handcrafted features and template-matching methods, which have limited flexibility and performance with high variability. In this work we investigate the use of Deep Metric Learning in the domain of activity recognition. We use a deep Triplet Network to generate fixed-length descriptors from activity samples for purposes of classification. We carry out evaluation of our proposed method on five datasets from different sources with differing activities. We obtain classification accuracies of up to 96% in self-testing scenarios and up to 91% accuracy in cross-dataset testing without retraining. We also show that our method performs similarly to traditional Convolutional Neural Networks. The obtained results indicate the promise of this approach.