Work

Human Fall Detection with Infrared Imaging: A Comparison of Graph Convolutional Networks and YOLO

Year: 2026

Type: article

Abstract: This paper presents a comparative study of two artificial intelligence approaches-graph convolutional networks (GCNs) and the YOLO object detection algorithm-for analyzing human fall events using infrared imaging. From the AI perspective, the study introduces a GCN model that achieves over 99% classification accuracy by modeling 2D and 3D skeletal data as graph structures and evaluates the real-time detection capabilities of YOLOv8 on infrared video frames. On the engineering side, the research addresses practical challenges in elderly care and healthcare monitoring systems by demonstrating how these AI methods can accurately detect and classify fall directions under infrared conditions. The results highlight each model's strengths and propose a hybrid framework combining YOLO's spatial localization with GCN's motion-pattern analysis for future real-world applications.

Source: Sensors

DOI: 10.3390/s26092794

Authors: Karol Perliński, Artur Faltyński, Aleksandra Świetlicka

Institution: Poznań University of Technology

Language: English

Topic: Human Pose and Action Recognition

Subfield: Computer Vision and Pattern Recognition

Field: Computer Science

Domain: Physical Sciences

SDG: Good health and well-being

Open Access status: gold

Estimated APC paid: $3,137