Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Nov 9;23(22):9069.
doi: 10.3390/s23229069.

A Lightweight Human Fall Detection Network

Xi Kan  1 , Shenghao Zhu  2 , Yonghong Zhang  1   2 , Chengshan Qian  1   2
Affiliations

A Lightweight Human Fall Detection Network

Xi Kan et al. Sensors (Basel). .

Abstract

The rising issue of an aging population has intensified the focus on the health concerns of the elderly. Among these concerns, falls have emerged as a predominant health threat for this demographic. The YOLOv5 family represents the forefront of techniques for human fall detection. However, this algorithm, although advanced, grapples with issues such as computational demands, challenges in hardware integration, and vulnerability to occlusions in the designated target group. To address these limitations, we introduce a pioneering lightweight approach named CGNS-YOLO for human fall detection. Our method incorporates both the GSConv module and the GDCN module to reconfigure the neck network of YOLOv5s. The objective behind this modification is to diminish the model size, curtail floating-point computations during feature channel fusion, and bolster feature extraction efficacy, thereby enhancing hardware adaptability. We also integrate a normalization-based attention module (NAM) into the framework, which concentrates on salient fall-related data and deemphasizes less pertinent information. This strategic refinement augments the algorithm's precision. By embedding the SCYLLA Intersection over Union (SIoU) loss function, our model benefits from faster convergence and heightened detection precision. We evaluated our model using the Multicam dataset and the Le2i Fall Detection dataset. Our findings indicate a 1.2% enhancement in detection accuracy compared with the conventional YOLOv5s framework. Notably, our model realized a 20.3% decrease in parameter tally and a 29.6% drop in floating-point operations. A comprehensive instance analysis and comparative assessments underscore the method's superiority and efficacy.

Keywords: GDCN module; GSConv module; NAM; SIoU; YOLOv5; fall detection.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1
Figure 1
The architecture of CGNS-YOLO.
Figure 2
Figure 2
Ordinary Ghost module.
Figure 3
Figure 3
Information flow diagram of DFC.
Figure 4
Figure 4
The architecture of the GDCN module.
Figure 5
Figure 5
The architecture of the GSConv module.
Figure 6
Figure 6
The architecture of the NAM module.
Figure 7
Figure 7
The scheme for calculating the contribution of the angle cost to the loss function.
Figure 8
Figure 8
Typical fall detection images.
Figure 9
Figure 9
Image division: Where (a) is the target center-of-mass location distribution, and (b) is the image size distribution.
Figure 10
Figure 10
Model training result curves: (a) is the train/box_loss curve, (b) is the train/cls_loss curve, (c) is the mAP0.5 curve, and (d) is the mAP0.5:0.95 curve.
Figure 11
Figure 11
Different light conditions.
Figure 12
Figure 12
Occlusion scenarios.
See this image and copyright information in PMC

References

    1. Delgado-Escano R., Castro F.M., Cozar J.R., Marin-Jimenez M.J., Guil N., Casilari E. A cross-dataset deep learning-based classifier for people fall detection and identification. Comput. Methods Programs Biomed. 2020;184:105265. doi: 10.1016/j.cmpb.2019.105265. - DOI - PubMed
    1. Girshick R. Fast R-CNN; Proceedings of the 2015 IEEE International Conference on Computer Vision (ICCV); Santiago, Chile. 7–13 December 2015; New York, NY, USA: IEEE; 2015. pp. 1440–1448.
    1. Dai J., Li Y., He K., Sun J. R-fcn: Object detection via region-based fully convolutional networks. Adv. Neural Inf. Process. Syst. 2016;29
    1. He K., Gkioxari G., Dollár P., Girshick R. Mask R-CNN; Proceedings of the 2017 IEEE International Conference on Computer Vision (ICCV); Venice, Italy. 22–29 October 2017; New York, NY, USA: IEEE; 2017. pp. 2961–2969.
    1. Cai Z., Vasconcelos N. Cascade R-CNN: Delving into high quality object detection; Proceedings of the 2018 IEEE/CVF Conference on Computer Vision And Pattern Recognition; Salt Lake City, UT, USA. 18–23 June 2018; New York, NY, USA: IEEE; 2018. pp. 6154–6162.