Ambient Diagnostics addresses innovative methods for discovering patterns from affordable devices, such as mobile phones, watches, cameras, and game interfaces, to interpret multimedia data for personal health monitoring and diagnosis. This is the first comprehensive textbookon multidisciplinary innovations in affordable healthcare—from sensory fusion, pattern detection, to classification. Connecting the Dots The material in this book combines sensing, pattern recognition, and visual design, and is divided into four parts, which cover fundamentals, multimedia intelligence, pervasive sensors, and crowdsourcing. The author describes basic pattern discovery models, sound, color, motion and video analytics, and pattern discovery from games and social networks. Each chapter contains the material’s main concepts, as well as case studies, and extensive study questions. Contains overviews about diagnostic sensors on mobile phones Reflects the rapidly growing platforms for remote sensing, gaming, and social networking Incorporates cognitive tests such as fatigue detection Includes pseudo code and sample code Provides vision algorithms and multimedia analytics Covers Multimedia Intelligence Extensively Ambient Diagnostics includes concepts for ambient technologies such as point-and-search, the pill camera, active sensing with Kinect, digital human labs, negative and relative feature spaces, and semantic representations. The book also introduces methods for collective intelligence from online video games and social media.

This book attempts to connect artificial intelligence to primitive intelligence. It explores the idea that a genuinely intelligent computer will be able to interact naturally with humans. To form this bridge, computers need the ability to recognize, understand and even have instincts similar to humans. The author organizes the book into three parts. He starts by describing primitive problem-solving, discussing topics like default mode, learning, tool-making, pheromones and foraging. Part two then explores behavioral models of instinctive cognition by looking at the perception of motion and event patterns, appearance and gesture, behavioral dynamics, figurative thinking, and creativity. The book concludes by exploring instinctive computing in modern cybernetics, including models of self-awareness, stealth, visual privacy, navigation, autonomy, and survivability. Instinctive Computing reflects upon systematic thinking for designing cyber-physical systems and it would be a stimulating reading for those who are interested in artificial intelligence, cybernetics, ethology, human-computer interaction, data science, computer science, security and privacy, social media, or autonomous robots.

A modern city generates a large volume of digital information, especially in the form of unstructured image and video data. Recent advancements in deep learning techniques have enabled effective learning and estimation of high-level attributes and meaningful features from large digital datasets of images and videos. In my thesis, I explore the potential of applying deep learning to image and video data to quantify urban tree cover and street parking utilization. Large-scale and accurate quantification of urban tree cover is important towards informing government agencies in their public greenery efforts, and useful for modelling and analyzing city ecology and urban heat island effects. We apply state-of-the-art deep learning models, and compare their performance to a previously established benchmark of an unsupervised method. Our training procedure for deep learning models is novel; we utilize the abundance of openly available and similarly labelled street-level image datasets to pre-train our model. We then perform additional training on a small training dataset consisting of GSV images. We also employ a recently developed method called gradient-weighted class activation map (Grad-CAM) to interpret the features learned by the end-to-end model. The results demonstrate that deep learning models are highly accurate, can be interpretable, and can also be efficient in terms of data-labelling effort and computational resources. Accurate parking quantification would inform developers and municipalities in space allocation and design, while real-time measurements would provide drivers and parking enforcement with information that saves time and resources. We propose an accurate and real-time video system for future Internet of Things (IoT) and smart cities applications. Using recent developments in deep convolutional neural networks (DCNNs) and a novel intelligent vehicle tracking filter, the proposed system combines information across multiple image frames in a video sequence to remove noise introduced by occlusions and detection failures. We demonstrate that the proposed system achieves higher accuracy than pure image-based instance segmentation, and is comparable in performance to industry benchmark systems that utilize more expensive sensors such as radar. Furthermore, the proposed system can be easily configured for deployment in different parking scenarios, and can provide spatial information beyond traditional binary occupancy statistics.

Ambient Intelligence refers to smart electronic environments that are sensitive and responsive to the presence of people. This book originates from the Workshop on Ambient Intelligence in Everyday Life held in San Sebastian, Spain, July 2005. Coverage is devoted to the cognitive aspects of ambient intelligence. The 15 carefully reviewed and revised articles presented are organized in topical sections on human-centric computing, ambient interfaces, and architectures for ambient intelligence.

Abstract: "The bandwidth of the wireless network has been a bottleneck for live security video streaming. In this paper, we introduce the 'Bandwidth of human visual processing' concept. We intend to match the human information processing bandwidth with physical network bandwidth. An eye gaze-based interface is used to optimize the video flow by adjusting the video resolution and compression. A face-based compression method allows enhancing the results to save the bandwidth up to 88%."

Metal-organic frameworks (or coordination polymers) are a recently-identified class of porous polymeric materials, consisting of metal ions or clusters linked together by organic bridging ligands. The major advantage of MOFs over other traditional materials, such as zeolites or activated carbons, is that their synthesis methods have provided an extensive class of crystalline materials with high stability, tunable metrics, and organic functionality. The ability to modify the physical environment of the pores and cavities within MOFs allow tuning of the interactions with guest species, and serves as a route to tailor the chemical stability and/or reactivity of the frameworks for specific applications. The classical way to incorporate functional groups into a MOF is the modification of the organic precursor with specific substituents before synthesizing the MOF itself; we call this approach pre-functionalization method. Functionalization of organic precursors is the initial and necessary step to obtaining functionalized isostructural MOFs and also provides the possibility for the post-synthetic modification of MOFs. However, in some cases, the functional groups may interfere with MOF synthesis and alter the topology of desired MOF. The goal of this proposed research is to explore the possibilities of metal-organic frameworks (MOFs) as novel porous structures, to study the effect of functional groups on the topologies and adsorption behavior of MOFs, and to understand how the synthesis conditions affect the phase purity and the in-situ reaction of ligands.

Abstract: "The public transportation is a cheap and ecological manner to travel in a city. However, the security in the buses and the subway is less efficient than in the airports. Checking all the bags in a subway and searching the passengers are the only ways to paralyze a city. The investigators propose a location based baggage detection for transit vehicles. The system looks for the high probability location of the baggage and classifies the warnings according to the stillness time of the baggage in the vehicle and the behavior of the baggage holder. A light location mask is introduced to enhance the accuracy of the detection."