From Offline to Real-Time Distributed Activity Recognition in Wireless Sensor Networks for Healthcare: A Review.

Area of Science:

• * Healthcare Technology
• * Artificial Intelligence
• * Signal Processing

Background:

• * Reviews foundational concepts in offline activity recognition, including sensor types, data labeling, feature extraction, outlier detection, and machine learning.
• * Discusses challenges in real-time centralized activity recognition, focusing on communication, labeling, cloud/local processing, and machine learning in streaming data.
• * Explores existing implementations of real-time distributed activity recognition in scientific literature.

Purpose of the Study:

• * To provide a comprehensive overview of the state-of-the-art in real-time distributed activity recognition for healthcare applications.
• * To identify and define key research challenges and optimization angles in this domain.
• * To serve as a guide for researchers and practitioners interested in distributed artificial intelligence and activity recognition.

Main Methods:

• * Comprehensive literature review of existing research in activity recognition.
• * Analysis of concepts from offline and real-time centralized activity recognition.
• * Identification and categorization of optimization strategies for real-time distributed systems.

Main Results:

• * Defined six primary optimization angles for real-time distributed activity recognition: Processing, memory, communication, energy, time, and accuracy.
• * Highlighted the transition from offline to real-time and distributed approaches, outlining specific challenges at each stage.
• * Synthesized current implementations and research trends in the field.

Conclusions:

• * Real-time distributed activity recognition presents significant challenges but offers substantial potential in healthcare.
• * Optimization across multiple dimensions (processing, memory, communication, energy, time, accuracy) is crucial for practical implementation.
• * The review provides a framework for future research and development in intelligent healthcare systems.