Development of weighted residual RNN model with hybrid heuristic algorithm for movement recognition framework in...

Area of Science:

• Healthcare Technology
• Artificial Intelligence
• Gerontology

Background:

• Ambient Assisted Living (AAL) systems are crucial for elderly and disabled individuals, offering safety and assistance.
• Movement recognition within AAL is an emerging field vital for ensuring independent and secure living.
• Current systems require advanced methods for accurate and reliable human movement detection.

Purpose of the Study:

• To develop an intelligent and automatic movement recognition system for Ambient Assisted Living (AAL) applications.
• To enhance healthcare assistance for elderly and disabled persons through accurate movement analysis.
• To propose a novel deep learning model integrated with a hybrid optimization algorithm for superior performance.

Main Methods:

• Utilized a Convolutional Autoencoder for extracting deep features from input data.
• Developed a Weighted Residual Recurrent Neural Network (RRNN) as the core movement recognition model.
• Employed a Hybrid Rat Swarm with Coati Optimization Algorithm to optimize RRNN model weights for improved training and testing.

Main Results:

• The proposed system demonstrated significant improvements in system performance and accuracy.
• Experimental validation confirmed the efficacy of the strategy in recognizing human movements.
• The optimized deep learning model achieved standard and reliable results.

Conclusions:

• The developed movement recognition system effectively supports Ambient Assisted Living goals.
• The integration of deep learning and hybrid optimization offers a promising approach for healthcare applications.
• The system's accuracy in movement recognition can significantly aid in providing timely medical assistance.