Adaptive motion assisted human activity recognition for people with disabilities via osprey optimisation-based dimensionality reduction with recurrent neural network

Key Points

• Human activity recognition showed significant improvement using osprey optimisation methods, enhancing recognition accuracy.
• The metrics indicated a high accuracy rate of 87% in recognizing various activities when utilizing the recurrent neural network approach.
• Analysis using osprey optimisation-based dimensionality reduction technique ensures effective data processing for increased output.
• This method may enable better technology integration for individuals with disabilities, paving the way for more inclusive support systems.