ABSTRACT

Over the past years, the world’s population has continued its transition from a state of high birth and death rates to the one characterized by low birth and death rates. A consequence of this transition is currently the aging population, which is putting severe stress on modern social and medical welfare systems [1]. One way of tackling the problem is to prolong the time for which persons live in their own homes independently by using assistive technologies. To this end, information and communication technologies assisting the elderly are of high social and economic relevance. Examples of assistive technologies are reminder systems, medical assistance and tele-healthcare systems, personal emergency response systems, social robotics, and accessible human-machine interfaces [2]. However, when deploying such systems in a home, several considerations must be taken into account. On the one hand, the development of ambient assisted living (AAL) systems usually relies on applicationdependent sensors, such as vital sensors, cameras, and microphones. On the other hand, since such systems are generally complex to use, the design of simple and intuitive interfaces is of great importance. It has been shown that natural and convenient ways to interact with technical systems are highly desired [3]. One way to develop an intuitive human-machine interaction is by using sound. The use of sound in AAL systems provides many advantages, since it has been shown that microphones can

9.1 Introduction .................................................................................................. 159 9.2 Wireless Acoustic Sensor Networks ............................................................. 162 9.3 Audio-Enabled Resource-Constrained Platforms ........................................ 164 9.4 Applications in Ambient Assisted Living ..................................................... 167 9.5 Sound Source Localization ........................................................................... 170 9.6 Acoustic Activity Detection .......................................................................... 173 9.7 Audio Event Detection .................................................................................. 175 9.8 Conclusion .................................................................................................... 176 References .............................................................................................................. 177

be easily integrated into existing living environments, they may serve multiple purposes (e.g., event detection, localization, and speech recognition), and users do not perceive them as obtrusive [4]. In fact, a monitoring device such as a video camera may be considered a severe break of privacy and, thus, is judged critically by the users. In contrast, acoustic monitoring can be considered a more accepted alternative. If the signal analysis is performed automatically by the system without storing acoustic data and without the possibility that other people may listen to the signals, this type of analysis would likely not be considered a break of privacy at all [5].