CoDEm: Conditional domain embeddings for scalable human activity recognition

• Main Authors: FARIDEE, Abu Zaher Md, CHAKMA, Avijoy, HASAN, Zahid, ROY, Nirmalya, MISRA, Archan
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2022
• Subjects: human activity recognition; domain embedding; attention; multi-task learning; adversarial learning; meta-data; Databases and Information Systems; Data Science; Health Information Technology
• Online Access: https://ink.library.smu.edu.sg/sis_research/7889; https://ink.library.smu.edu.sg/context/sis_research/article/8889/viewcontent/8._Zaher_Smartcomp2022_CoDEm.pdf
• Institution: Singapore Management University

We explore the effect of auxiliary labels in improving the classification accuracy of wearable sensor-based human activity recognition (HAR) systems, which are primarily trained with the supervision of the activity labels (e.g. running, walking, jumping). Supplemental meta-data are often available d...