the Internet of things (IoT) can be understood as a network of smart devices connected to the Internet that collect and share data. the Internet of Healththings (IoHT) is an area of research that has been gaining a l...
the Internet of things (IoT) can be understood as a network of smart devices connected to the Internet that collect and share data. the Internet of Healththings (IoHT) is an area of research that has been gaining a lot of prominences and includes the use of IoT solutions aimed at monitoring healthcare and improving the health of users of these solutions. the development of IoHT solutions involves several challenges, including the interoperability between different smart devices and different sets of sensors, the difficulties of initial design decisions about which technology to use for the solution, and the development cost for the smart devices, which includes not only the financial cost but also the processing and energetic cost that are important to be considered, since the hardware limitations of these smart devices can impede the execution of some applications. the reuse of software artifacts can help reduce costs and other challenges. In this sense, this work proposes the modeling and implementation of a classification graph that relates different sensors, features, classification algorithms, and health states (or situations), providing a reusable software artifact that can help boththe requirements elicitation and design stages. Also, the classification graph can be used as a knowledge base for implementing data analysis modules and predicting health states. We evaluated the proposal through a proof of concept, in which we implemented a classification graph based on the proposed model in a web server, using two different datasets, with data from accelerometer and gyroscope sensors and 30 actions (or final states). the developed system also implements three classification algorithms: an artificial neural network, a decision tree, and a random forest. Moreover, we developed a Web API to execute requests for boththe creation and update of the classification graph and the request to download the optimized graph and the trained models by the classification algo
Haze, a phenomenon arising from the scattering of light by atmospheric particles, imparts undesirable visual artifacts and degradation to images, posing challenges across various applications. this review paper delves...
Haze, a phenomenon arising from the scattering of light by atmospheric particles, imparts undesirable visual artifacts and degradation to images, posing challenges across various applications. this review paper delves into the realm of dehazing algorithms, with a specific focus on their hardware implementations. the introductory section lays the groundwork by elucidating the nature of haze, its diverse types, and the detrimental impacts it inflicts on image quality through light scattering mechanisms. In response to the pressing need for real-time dehazing solutions, the paper navigates through the landscape of hardware-based dehazing algorithms this review addresses the intricate hurdles encountered in translating dehazing algorithms into hardware architectures. these challenges span architectural design considerations, power efficiency optimizations, memory bandwidth management, and algorithmic adaptation for parallel processing. To benchmark the advancements achieved thus far, the paper delves into a comparative analysis of diverse hardware implementations and their corresponding results. Notable architectural choices, from FPGA (Field-Programmable Gate Array) to specialized ASICs (Application-Specific Integrated Circuits) are dissected in terms of their performance metrics, energy efficiency, and suitability for real-time applications. this assessment also encompasses an examination of trade-offs between accuracy and computational speed in various hardware contexts. Drawing insights from the reviewed literature, this paper offers a comprehensive overview of the strides taken in the field of hardware-based dehazing algorithms. the synthesis of challenges, architectures, and comparative results provides a roadmap for future research directions, offering insights into potential areas of improvement and innovation.
Perceiving the position and orientation of objects (i.e., pose estimation) is a crucial prerequisite for robots acting within their natural environment. We present a hardware acceleration approach to enable real-time ...
详细信息
Perceiving the position and orientation of objects (i.e., pose estimation) is a crucial prerequisite for robots acting within their natural environment. We present a hardware acceleration approach to enable real-time and energy efficient articulated pose estimation for robots operating in unstructured environments. Our hardware accelerator implements Nonparametric Belief Propagation (NBP) to infer the belief distribution of articulated object poses. Our approach is on average, 26X more energy efficient than a high-end GPU and 11X faster than an embedded low-power GPU implementation. Moreover, we present a Monte-Carlo Perception Library generated from high-level synthesis to enable reconfigurable hardware designs on FPGA fabrics that are better tuned to user-specified scene, resource, and performance constraints.
Modular converters gained popularity in low to high voltage applications since it was first introduced in early 2000s due to the design flexibility. Conventional modular converters require large submodule capacitance ...
Modular converters gained popularity in low to high voltage applications since it was first introduced in early 2000s due to the design flexibility. Conventional modular converters require large submodule capacitance to accommodate single-phase AC processing power requirement that lead to the use of electrolytic capacitors. the use of electrolytic capacitor in each submodule increases the converter volume and affects the efficiency of the converter. this paper presents an integrated modular topology with integrated positive and negative arms and integrated half-bridge modules. the proposed circuit design permits significant reduction in energy storage requirement due to energy balance per switching cycle and thus a tiny shared capacitor (few micro-farads versus several milli-farads). the number of capacitors required in the converter is also reduced significantly because only one capacitor is needed between the positive and negative arms. the work is verified experimentally using a laboratory-scale prototype and circuit simulations.
Readout integrated circuits (RIOC) play a vital role in interpreting meaningful results with high precision for physico-chemical models like Ion-sensitive FET. this paper presents a design and analysis of three compos...
详细信息
ISBN:
(纸本)9781728158754
Readout integrated circuits (RIOC) play a vital role in interpreting meaningful results with high precision for physico-chemical models like Ion-sensitive FET. this paper presents a design and analysis of three composite based readout circuits for ion-sensitive field effect transistor (ISFET) and Al2O3 sensing film. Two improved readout circuit of 0.5 mu m technology have been simulated in LTspice XVII software. A macro model of ISFET has been used in all three circuits to check and test relationship between I-d and V-d, V-ref and V-pH. Relationship between I-d and V-ref has also been checked at different pH values to observe the behavior of ISFET for the FVFCS MOSFET readout circuit. All the three developed circuits have been briefly compared through transfer and output characteristic graphs. Simulation results infer average sensitivity value achieved up-to 51.5 mV/pH for proposed FVFCS MOSFET readout circuit. All the three circuits exhibit a linear variation in drain current (I-d) and equivalent voltage pH (V-pH) values.
In this work, we consider a downlink non-orthogonal multiple access (NOMA) network where multiple single-antenna users are served by multiple multi-antenna base stations (BSs). For practical considerations, we assume ...
详细信息
ISBN:
(纸本)9783031231407;9783031231414
In this work, we consider a downlink non-orthogonal multiple access (NOMA) network where multiple single-antenna users are served by multiple multi-antenna base stations (BSs). For practical considerations, we assume that only the imperfect channel state information (CSI) of each user is available at the BSs. Based on this model, the problem of joint user grouping and robust beamforming design is formulated to minimize the sum transmission power, and meanwhile, guarantee the quality of service requirements of users. Due to the integer variables of user grouping, coupling effects of beamformers, and infinitely many constraints caused by the imperfect CSI, the formulated problem is challenging to solve. For computational complexity reduction, the original problem is divided into a user grouping subproblem and a robust beamforming design subproblem. First, the user grouping problem is efficiently solved by a coalition formation game based algorithm. then, for the robust beamforming design problem, a semidefinite relaxation (SDR) based method is proposed to produce a suboptimal solution efficiently. Moreover, we provide a sufficient condition under which the SDR based approach can guarantee to obtain an optimal rank-one solution, which is theoretically analyzed. Simulation results demonstrate the efficacy of the proposed algorithms.
Mobile Science Center is a Polish project that seeks to bring astronomy knowledge to wider social groups through various applications. In its development it is necessary to design a graphical interface that explains a...
详细信息
Historically, we are taught to use task-dependent architecture design and objectives to tackle data science tasks. Counter intuitively, this dogma has been proven (partly) wrong by deep neural networks, while rising i...
详细信息
ISBN:
(数字)9798350364941
ISBN:
(纸本)9798350364958
Historically, we are taught to use task-dependent architecture design and objectives to tackle data science tasks. Counter intuitively, this dogma has been proven (partly) wrong by deep neural networks, while rising into stardom, e.g., large video/language models, widely used for natural language, image, speech, and other multi-modal tasks with unprecedented performance across data science. Key attributes of current foundation models is their ability to fuse different information (potentially across different domains, e.g., text and images), and their ability to capture high-order interactions of the elements using the Transformer architecture [1]. In this tutorial, we will focus on architecture design and explain how the transition from older ideas of task-dependent architectures to large, pretrained foundation models emerged. We will also identify fundamental flaws in the computational efficiency and study the impact of architectural choices into generalization, robustness, privacy, and fairness.
In this work, we propose to predict four items from the UCLA loneliness scale using subject self-report scores. Using subjective self-reporting, over 14 days, for positive and negative affect, and depression and anxie...
In this work, we propose to predict four items from the UCLA loneliness scale using subject self-report scores. Using subjective self-reporting, over 14 days, for positive and negative affect, and depression and anxiety we evaluate both subject dependent (personalized) and subject independent experimental design. We evaluate four approaches for prediction, namely random forest, support vector machine, k-nearest neighbor, and logistic regression. We find that the features (self-report) are relatively stable across all four approaches. Along with each individual self-report feature, we also evaluate the fusion of all features where they are concatenated into one feature vector. through our experimental design, we show that UCLA loneliness items can be predicted, and that the fusion of features (positive and negative affect, and depression and anxiety) is the most encouraging way to do this prediction. We also show that the number of days, used for prediction, has a noticeable impact on the results and that personalization helps with prediction.
Tailoring the training process according to recovery potentials has gained importance in the process of training. Nowadays, the intelligent hospital is coming into sight, and the traditional rehabilitation assessment ...
详细信息
ISBN:
(纸本)9783030991913;9783030991906
Tailoring the training process according to recovery potentials has gained importance in the process of training. Nowadays, the intelligent hospital is coming into sight, and the traditional rehabilitation assessment has been unable to meet the development of the times. In order to meet the demand, a dynamic assessment of the performance of the recovery process is required. Cloud computing can calculate and store massive data, and its application in the evaluation system of limb motor function using inertial sensor can meet the requirements of hospitals for data security, resource sharing, maintainability and so on. In order to accurately assess rehabilitation for the upper limb, the inertial sensors are used to collect the real-time limb movement data of patients. In the next steps, an evaluation of the patients rehabilitation exercises results is presented, saved and statistically analyzed. the assessment can help achieve more individualized patient care. therefore, this paper designs an evaluation system of limb motor function using inertial sensor, which effectively improves the efficiency of rehabilitation assessment. the evaluation system increases the real-time action display on the screen to improve the practicability of the evaluation system. At present, the evaluation system is in the development stage, and a lot of data and work are still needed to improve the evaluation system.
暂无评论