Home About Login Current Archives Announcements Editorial Board
Submit Now For Authors Call for Submissions Statistics Contact
Home > Archives > Volume 20, No 15 (2022) > Article

DOI: 10.14704/NQ.2022.20.15.NQ88010

Elderly Care Monitoring using IOT and Deep Learning

Dr. Reena Singh, Poonam J. Patil, Dr. B.K. Sarkar

Abstract

As of late, the strategies of Web of Things (IoT) and versatile interchanges have been created to assemble human and climate data information for different clever administrations and applications. Remote observing of old and crippled individuals residing in shrewd homes is profoundly difficult because of likely mishaps which could happen because of everyday exercises like falls. For older individuals, fall is considered as a significant justification behind death of post-horrible intricacy. In this way, early recognizable proof of old individuals falls in shrewd homes is expected to build the endurance pace of the individual or deal needed help. As of late, the approach of computerized reasoning (man-made intelligence), IoT, wearables, cell phones, and so on makes it practical to configuration fall location frameworks for shrewd homecare. In this view, this paper presents an IoT empowered older fall discovery model utilizing ideal profound convolutional brain organization (IMEFD-ODCNN) for brilliant homecare. The objective of the IMEFD-ODCNN model is to empower cell phones and wise profound learning (DL) calculations to distinguish the event of falls in the shrewd home. Principally, the info video caught by the IoT gadgets is pre-handled in various ways like resizing, expansion, and min-max based standardization. Moreover, Crush Net model is utilized as an element extraction procedure to infer proper component vectors for fall discovery. What's more, the hyperparameter tuning of the SqueezeNet model happens utilizing the salp swarm improvement (SSO) calculation. At last, sparrow inquiry improvement calculation (SSOA) with variational autoencoder (VAE), called SSOA-VAE based classifier is utilized for the arrangement of fall and non-fall occasions. At long last, if there should be an occurrence of fall occasion identified, the cell phone sends an alarm to the overseers and medical clinic the executives. The presentation approval of the IMEFD-ODCNN model happens on UR fall discovery dataset and numerous cameras fall dataset. The exploratory results featured the promising execution of the IMEFD-ODCNN model over the new techniques with the most extreme precision of 99.76% and 99.57% on the numerous cameras fall and UR fall identification dataset.

Keywords

Hyperparameter, Tuning,Smart homecare, Smartphone, Fall Detection, Artificial intelligence, Elderly people, Deep learning, Parameter tuning

Full Text

PDF

References

?>