COMPUTER VISION FOR ASSISTIVE HEALTHCARE PREFACE

This book reflects the importance of considering the "real world" social impact of technology alongside the fundamental goals of basic R& D. Computer vision is an area of scientific and technological development that will continue to have a profound impact on society. It will redefine the way that information technology intersects and interfaces with medicine and other disciplines, and it will play a key role in the care of an aging population and in improving the quality of life in our modern society. The main aim of this book is to present the state of the art in the context of Computer Vision for Assistive Healthcare. The different chapters present the latest progress in this domain and discuss novel ideas in the field. In addition to the technologies used, emphasis is given to the definition of the problems, the available benchmark databases, the evaluation protocols, and procedures. Chapter 1, by Zhigang Zhu et al., presents a vision-based assistive indoor localization approach with various techniques in different stages for helping the visually impaired to localize in and navigate through indoor environments. Unlike other computer vision research, whose problems are already well-defined and formalized by the community and whose major tasks are to apply their developed algorithms to standard datasets by tuning the parameter of models and evaluating the performance, this work studies the navigation needs of the visually impaired, and then helps us develop techniques in data collection, model building, localization, and user interfaces in both pre-journey planning and real-time assistance. Chapter 2, by Corneliu Florea et al., approaches computer vision solutions to the diagnostic aid of several cognitive-affective psychiatric disorders. It reviews contributions that investigate cognitive impairments that appear at all stages of human development: from childhood and cognitive accumulation (autism, dyslexia), through adulthood and trauma-related cognitive degradations (such as phobias and PTSD), and ending with the ultimate degenerative cognitive degradations induced by dementia. Chapter 3, by Antonio Frisoli et al., describes a computer vision-based robot-assisted system used in neurorehabilitation of post-stroke patients, which allows the subjects to reach and grasp objects in a defined workspace. In particular, a novel RGB-D-based algorithm used to track generic unknown objects in real time is proposed. The novelty of the proposed tracking algorithm comes from combining different features to achieve object recognition and tracking Chapter 4, by Nassir Navab et al., outlines how computer vision can support the surgeon during an intervention using the example of surgical instrument tracking in retinal microsurgery, which incorporates challenges and requirements that are common when using this technique in various medical applications. In particular, how to derive algorithms for simultaneous tool tracking and pose estimation based on random forests and how to increase robustness to problems associated with retinal microsurgery images, such as strong illumination variations and high noise levels, is shown. Chapter 5, by Qiuhong Ke et al., focuses on the gesture recognition task for HMI and introduces current deep learning methods that have been used for human motion analysis and RGB-D-based gesture recognition. More specifically, it briefly introduces the convolutional neural networks (CNNs), and then presents several deep learning frameworks based on CNNs that have been used for gesture recognition by using RGB, depth, and skeleton sequences. Chapter 6, by Sara Colantonio et al., offers a brief survey of existing, vision-based monitoring solutions for personalized health care and wellness, and introduces the Wize Mirror, a multisensory platform featuring advanced algorithms for cardio-metabolic risk prevention and quality-of-life improvement. Chapter 7, by Mariella Dimiccoli, focuses on those aspects of egocentric vision that can be directly exploited to develop platforms for ubiquitous context-aware personal assistance and health monitoring, also highlighting potential applications and further research opportunities in the context of assistive technologies. Chapter 8, by Sethuraman Panchanathan et al., presents the computer vision research contributions of the Center for Cognitive Ubiquitous Computing (CUbiC) at Arizona State University in the design and development of a Social Interaction Assistant (SIA), which is an Augmentative and Alternative Communication (AAC) technology that can enrich the communication experience of individuals with visual impairment. The proposed solutions place emphasis on understanding the individual user's needs, expectations, and adaptations towards designing, developing, and deploying effective multimedia solutions. Empirical results demonstrate the significant potential in using person-centered AAC technology to enrich the communication experience of individuals with visual impairments. Chapter 9, by Peng Wang et al., focuses on the most recent research methods in understanding visual lifelogs, including semantic annotations of visual concepts, utilization of contextual semantics, recognition of activities, and visualization of activities. Some research challenges which indicate potential directions for future research are also discussed. Chapter 10, by Oya Celiktutan et al., focuses on recent advances in social robots that are capable of sensing their users, and support their users through social interactions, with the ultimate goal of fostering their cognitive and socio-emotional well-being. This chapter sets out to explore automatic analyses of social phenomena that are commonly studied in the fields of affective computing and social signal processing, together with an overview of recent vision-based approaches used by social robots. The chapter then describes two case studies that demonstrate how emotions and personality, which are two key phenomena for enabling effective and engaging interactions with robots, can be automatically predicted from visual cues during human-robot interactions. The chapter concludes by summarizing the open problems in the field and discussing potential future directions. Chapter 11, by Andrea Cavallaro et al., covers models and algorithms for the analysis of crowds captured in videos that can facilitate personal mobility, safety, and security and can enable assistive robotics in public spaces. The main challenges and solutions for the analysis of collective behavior in public spaces are discussed; these challenges include understanding how people interact and their constantly changing interpersonal relations under clutter and frequent visual occlusions. Finally, Chapter 12, by Manuela Chessa et al., considers assistive environments and discusses the possible benefits for an aging population. As a study case the current state of research on a protected discharge model adopted by Galliera hospital (Genova, Italy) to assist elderly users after they have been dismissed from the hospital and before they are ready to go back home, with the perspective of coaching them towards a healthy lifestyle, are discussed. The chapter focuses in particular on the vision-based modules designed to automatically estimate a frailty index of the patient, which allows a physician to assess the patient's health status and state of mind.