Table 9 Comparison of proposed methodology with existing work. Results are for reference only due to differences in datasets, classes, and settings).

Na Zhu et al.³⁹

Deep Neural Network

4

86%

2021

Wenfeng Pang et al.⁴⁰

Convolutional Neural Network

2 (Fall/Non-Fall)

81.1%

2022

Jorge D. Cardenas et al.⁴¹

CNN and LSTM

4 (No-Activity, Walking, Up/Down Stairs and Falling)

92.1%

2023

Navdeep Kaur et al.⁴²

Hybrid Haar Cascade Model

2 (Normal/Fall)

89.2%

2024

Proposed Method

Deep Learning and MFO

4 (Forward-Fall, Back-Fall, Side-Fall, Non-Fall)

93.2%