We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Automatic Unstructured Handwashing Recognition using Smartwatch to Reduce Contact Transmission of Pathogens

Abstract: Current guidelines from the World Health Organization indicate that the SARSCoV-2 coronavirus, which results in the novel coronavirus disease (COVID-19), is transmitted through respiratory droplets or by contact. Contact transmission occurs when contaminated hands touch the mucous membrane of the mouth, nose, or eyes. Moreover, pathogens can also be transferred from one surface to another by contaminated hands, which facilitates transmission by indirect contact. Consequently, hands hygiene is extremely important to prevent the spread of the SARSCoV-2 virus. Additionally, hand washing and/or hand rubbing disrupts also the transmission of other viruses and bacteria that cause common colds, flu and pneumonia, thereby reducing the overall disease burden. The vast proliferation of wearable devices, such as smartwatches, containing acceleration, rotation, magnetic field sensors, etc., together with the modern technologies of artificial intelligence, such as machine learning and more recently deep-learning, allow the development of accurate applications for recognition and classification of human activities such as: walking, climbing stairs, running, clapping, sitting, sleeping, etc. In this work we evaluate the feasibility of an automatic system, based on current smartwatches, which is able to recognize when a subject is washing or rubbing its hands, in order to monitor parameters such as frequency and duration, and to evaluate the effectiveness of the gesture. Our preliminary results show a classification accuracy of about 95% and of about 94% for respectively deep and standard learning techniques.
Comments: 10 pages
Subjects: Machine Learning (cs.LG); Human-Computer Interaction (cs.HC); Signal Processing (eess.SP)
Cite as: arXiv:2107.13405 [cs.LG]
  (or arXiv:2107.13405v2 [cs.LG] for this version)

Submission history

From: Emanuele Lattanzi PhD. [view email]
[v1] Wed, 28 Jul 2021 14:52:45 GMT (389kb,D)
[v2] Mon, 15 Nov 2021 16:55:31 GMT (1159kb,D)

Link back to: arXiv, form interface, contact.