Enhancing indoor activity recognition for disabled persons using multi head self attention recurrent neural network with improved pelican algorithm.

Indoor activity monitoring methods ensure the well-being and security of elderly and visually impaired people living in their houses. These systems utilize numerous technologies and sensors to monitor day-to-day activities, including medication adherence, movement, and sleep patterns, providing a comprehensive view of the user's overall health and daily life. The accuracy and adaptability of deep learning (DL) models make human activity recognition (HAR) a valuable tool for enhancing effectiveness, security, and personalized experiences in indoor environments. HAR, utilizing DL techniques, innovates indoor monitoring by enabling precise understanding and detection of human actions. Deep neural networks (DNNs) analyze data from multiple sensors, such as cameras or accelerometers, to distinguish between various action patterns. DL models mechanically remove and learn discriminating features, making them suitable for identifying intricate human activities in sensor data. Nevertheless, selecting the appropriate DL architectures and optimizing their parameters is crucial for achieving improved solutions. This study proposes an Improved Pelican Optimisation for Indoor Activity Recognition in Persons with Disabilities using the Recurrent Neural Network (IPOIAR-DPRNN) method. The primary aim of the IPOIAR-DPRNN method is to enhance indoor activity detection systems for individuals with disabilities. Initially, the image pre-processing stage applies adaptive bilateral filtering (ABF) to reduce unwanted distortions or artefacts in the image. Furthermore, the EfficientNetB7 method is employed for the feature extraction process. For the detection and classification of indoor activities, the bidirectional long short-term memory with multi-head self-attention (BiLSTM-MHSA) technique is used. Additionally, the improved pelican optimization algorithm (IPOA)-based hyperparameter tuning is performed to enhance the detection results of the BiLSTM-MHSA technique. The validation of the IPOIAR-DPRNN approach is examined using the Florence 3D Actions dataset, and the outcomes are measured against various metrics. The comparison study of the IPOIAR-DPRNN approach revealed a superior accuracy value of 97.11% compared to existing techniques.