Multi-task Self-Supervised Learning for Human Activity Detection
Multi-task Self-Supervised Learning for Human Activity Detection is a scholarly work, published in 2019 in ''Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies''. The main subjects of the publication include labeled data, task, activity recognition, pattern recognition, feature learning, semi-supervised learning, supervised learning, computer science, deep learning, artificial intelligence, convolutional neural network, unsupervised learning, machine learning, leverage, anomaly detection, transfer of learning, and binary classification. The authors propose a novel self-supervised technique for feature learning from sensory data that does not require access to any form of semantic labels, i.e., activity classes.