Search

Picosecond ultrasonic laser sonar (PULSE), one of the noninvasive technologies with a high thickness sensitivity, is capable of probing multilayer films directly on the production wafers. To reach ...

In recent years, face detection has been widely applied to a variety of fields, such as face recognition, image focusing, and surveillance systems. This study proposes a real-time multi-faces detection system based on naive Bayesian classifier using Field Programmable Gate Array (FPGA). The system includes three main parts, feature extraction, candidate face detection, and false elimination. First, downscale the image to the image pyramid and extract local binary image features from each downscaling image. With the bit-plane slicing for Local Binary Pattern (LBP) can save the memory consumption and speed up the computation. Then, adopt the naive Bayesian classifier to identify candidate faces. Finally use skin color filter and face overlapping elimination to remove false positives. The experimental results show that the accuracy rate is up to 96.14% in face detection, which demonstrates the proposed real-time multi-faces detection system is indeed effective and efficient.

Spinel structured aluminates TAl

Thanks to advances in technology, the forth industrial revolution (Industry 4.0) is coming. All the manufacturers are undertaking large-scale technological innovations which include artificial intelligence and auto guided vehicle. These two research fields are also the major techniques in the proposed paper. In this paper, the tracking control and slope-climbing strategy between autonomous mobile robot and flatbed vehicle is proposed. The structure is integrated by three parts. First, the Q-learning algorithm is applied to controller design. Second, LIDAR sensor and camera are used to measure distance, forward direction and position of flatbed vehicle relative to autonomous mobile robot. Third, Robot Operation System (ROS) is adopt to be the data communication system among central processor unit, LIDAR sensor and camera of the autonomous mobile robot. In the simulation results, the flatbed vehicle follows three different trajectories, and the autonomous mobile robot computes tracking paths by machine vision and Q-learning algorithm. After reaching a certain distance, the autonomous mobile robot would carry out slope-climbing strategy to link with flatbed vehicle successfully.

Objectives: Previous studies have demonstrated that patients with diabetic retinopathy (DR) have a higher prevalence of risk factors known to be associated with cardiovascular disease (CVD). We hypothesized that patients with more severe DR could have a higher relative risk of CVD. Methods: To test this hypothesis, we used the National Health Insurance Research Database (NHIRD) to evaluate whether associations exist between DR and CVD. The data for this nationwide population-based retrospective cohort study were obtained from the NHIRD in Taiwan from 2001 to 2013. The assessed study outcome used was the incidence and other statistical analyses of CVD in patients with DR during a 13-year follow-up period. Results: Our findings obtained from 2001 to 2013 suggest that the incidence rates of CVD are 2.026 times that of diabetes mellitus (DM) without DR (95% C.I. = 1.876–2.187) and 2.75 times that of DM with DR (95% C.I. = 2.487–3.04) compared with the Non-DM group. Conclusion: The relative risk of CVD in DR was greater than that in the Non-DM group for both men and women. Targeted monitoring of DM, especially the co-existence of diabetic retinopathy, is of utmost importance in the clinical care of the DM population.

In photovoltaic (PV) systems, PV modules are usually cascaded in series to generate higher output power, and the P-V characteristic under partial shading condition (PSC) presents multiple peaks with nonlinearity. Several traditional tracking methods such as Perturbation and Observation (P&O) and Incremental Conductance (InC) cannot fulfill the global maximum power point (GMPP) tracing under PSC. Moreover, even the global maximum power point tracking (GMPPT) methods proposed recently such as Particle Swarm Optimization (PSO) may cost lots of time to locate the GMPP of the PV array. As a result, to ensure the PV array under PSC can still generate maximum power efficiently, this paper proposed the Adaptive Neuro Fuzzy Inference System (ANFIS)-based MPPT method, which apples ANFIS and the concept of particles to the MPPT without using expensive irradiance sensors. Instead of searching for the GMPP online, the proposed ANFIS-based MPPT use the training data pairs generated offline to economize the searching time and finish the GMPP tracing simultaneously with zero oscillation under simulation. In addition, it is shown that the tracking time and efficiency of the proposed MPPT method can reach average 6.04 steps and 99.85%, respectively, under simulation, and verifies the fast tracking speed under implementation. Compared to other methods, the proposed MPPT have faster tracking with higher efficiency and zero oscillation around the GMPP.

The maximum power point tracking (MPPT) technique is often used in photovoltaic (PV) systems to extract the maximum power in various environmental conditions. The perturbation and observation (P&amp, O) method is one of the most well-known MPPT methods, however, it may face problems of large oscillations around maximum power point (MPP) or low-tracking efficiency. In this paper, two reinforcement learning-based maximum power point tracking (RL MPPT) methods are proposed by the use of the Q-learning algorithm. One constructs the Q-table and the other adopts the Q-network. These two proposed methods do not require the information of an actual PV module in advance and can track the MPP through offline training in two phases, the learning phase and the tracking phase. From the experimental results, both the reinforcement learning-based Q-table maximum power point tracking (RL-QT MPPT) and the reinforcement learning-based Q-network maximum power point tracking (RL-QN MPPT) methods have smaller ripples and faster tracking speeds when compared with the P&amp, O method. In addition, for these two proposed methods, the RL-QT MPPT method performs with smaller oscillation and the RL-QN MPPT method achieves higher average power.

Face detection and recognition is an important issue and a difficult task in computer vision and human-computer interaction. Recently, with the development of deep learning, several related technologies have been proposed for face detection and facial expression recognition (FER), and the outstanding convolutional neural networks are the most common used in this field. This thesis applies the multi-task cascade convolutional neural network to face detection, and then designs the real-time FER system based on densely connected convolution network (DenseNet). The system first scales the input image to an image pyramid, and then uses the hierarchical network to determine whether a candidate window includes a human face. If a face exists, then send the candidate window to the FER system. Since DenseNet possesses the property of feature reuse, it can effectively reduce the amount of parameters and computation efforts, beneficial to develop the real-time system. In order to capture the variation of facial muscle in different expressions, this architecture adopts convolution operations with a stride 1 and tries different numbers of dense blocks. Through experiments, the proposed system can achieve real-time recognition in 30FPS and with recognition accuracy better than human eyes.

In this study, numerical calculations using single- and two-phase models of CuO/water nanofluid forced convection in a three-dimensional C-shaped channel with constant heat flux are investigated. The laminar heat transfer enhancement using a nanofluid in a chaotic flow is first validated with the available data in the literature and the maximum discrepancy is within 3%; then further it is extended to design the C-shaped geometry. In addition, after comparisons of the numerical results with single- and two-phase models, the multiparameter constrained the optimization procedure integrating the design of experiments (DOE), response surface methodology (RSM), genetic algorithm (GA), and computational fluid dynamics (CFD) is proposed to design the nanofluid laminar convection of three-dimensional C-shaped channels. The thermal performance factors predicted by the regression function for the C-shaped channel case are in good agreement with the numerical results of CFD, with the difference being within 10%.

Through the femtosecond-time-resolved study of photoacoustic pulse generation in ultra-thin gold nanofilms, we observed a time delay of 0.5–0.7 ps in the formation of thermal-expansion pulses after photoexcitation. Our observation indicates that lattice anharmonicity dominates over hot electron pressure in the thermal expansion of Au nanofilms under ultrashort-pulsed photoexcitation.

We investigate the effects of surface tunneling on electrostatics and transport properties of two-dimensional electron gases (2DEGs) in undoped Si/SiGe heterostructures with different 2DEG depths. By varying the gate voltage, four stages of density-mobility dependence are identified with two density saturation regimes observed, which confirms that the system transitions between equilibrium and nonequilibrium. Mobility is enhanced with an increasing density at low biases and, counterintuitively, with a decreasing density at high biases as well. The density saturation and mobility enhancement can be semiquantitatively explained by a surface tunneling model in combination with a bilayer screening theory.

In recent years, face detection is widely used in various fields, such as face recognition, image focusing, and surveillance systems. This study proposes a real-time face detection system based on naive Bayesian classifier using Field-programmable gate array(FPGA). The detection system divided into three main parts, feature extraction, candidate face detection, and false elimination. First, downscale the image to the image pyramid and extract local binary image features from each downscaling image; then features go through the naive Bayesian classifier to identify candidate faces. Finally, use skin color filter and face overlapping elimination to remove false positives. Detection results output to the monitor in VGA. In this paper, face detection system to implement in FPGA. As a result of the FPGA parallel processing, in 640×480 resolutions, the face detection of an image executes within 16.7 milliseconds; the improved local binary features, compared to Haar features, save around 140 times the amount of memory. The experimental results show that the accuracy rate is higher than 95% in face detection, which implies the proposed real-time detection system is indeed effective and efficient.

In recent years, the IoT application and the biometric-based authorization become popular. This paper proposes a face recognition system with high accuracy rate based on extended Local Binary Pattern, and applies it as an access control system on an IoT device which is always low-cost, low-power and small-footprint. The proposed face recognition system includes three parts, face detection, feature extraction and face recognition. For the face detection, the Viola-Jones face detector is adopted to find out the face information. The extended Local Binary Pattern then extracts the local features of the face. Further transform these features to a low-dimension subspace by Principle Component Analysis method. Finally, use the classification based on the sparse representation of L2 norm minimization to identify and verify the face. From the experimental results, the proposed method can achieve a high recognition rate better than 95% in several face databases, even reach 99% for the Cohn-Kanade face database. The access control system implemented on Raspberry Pi 3 is able to complete the whole face recognition in a second, which makes it indeed a real-time system.

We investigate DC characteristics of a two-dimensional electron gas (2DEG) in an undoped Si/SiGe heterostructure and its temperature dependence. An insulated-gate field-effect transistor was fabricated, and transfer characteristics were measured at 4 K–300 K. At low temperatures (T 45 K, the current saturation and negative transconductance disappear and the device acts as a normal transistor.

Heme oxygenase-1 (HO-1), a stress-inducible enzyme anchored in the endoplasmic reticulum (ER) by a single transmembrane segment (TMS) located at the C terminus, interacts with NADPH cytochrome P450 reductase and biliverdin reductase to catalyze heme degradation to biliverdin and its metabolite, bilirubin. Previous studies suggested that HO-1 functions as a monomer. Using chemical cross-linking, co-immunoprecipitation, and fluorescence resonance energy transfer (FRET) experiments, here we showed that HO-1 forms dimers/oligomers in the ER. However, oligomerization was not observed with a truncated HO-1 lacking the C-terminal TMS (amino acids 266-285), which exhibited cytosolic and nuclear localization, indicating that the TMS is essential for the self-assembly of HO-1 in the ER. To identify the interface involved in the TMS-TMS interaction, residue Trp-270, predicted by molecular modeling as a potential interfacial residue of TMS alpha-helices, was mutated, and the effects on protein subcellular localization and activity assessed. The results showed that the W270A mutant was present exclusively in the ER and formed oligomers with similar activity to those of the wild type HO-1. Interestingly, the W270N mutant was localized not only in the ER, but also in the cytosol and nucleus, suggesting it is susceptible to proteolytic cleavage. Moreover, the microsomal HO activity of the W270N mutant was significantly lower than that of the wild type. The W270N mutation appears to interfere with the oligomeric state, as revealed by a lower FRET efficiency. Collectively, these data suggest that oligomerization, driven by TMS-TMS interactions, is crucial for the stabilization and function of HO-1 in the ER.

This paper develops an Android face recognition application for users on mobile device, and applies it in the face verification system of Samsung Galaxy SII smart phone. The developed face recognition application includes three parts, the face detection using the Viola-Jones face detection program, the feature extraction implemented by the eigenface features, and the face recognition based on the sparse representation of L2 norm minimization. Different to general learning methods, the developed application does not require a tremendous amount of time in data training and can achieve a high recognition rate even higher than 99%, for examples 99.2% for the Sheffield face database, 99.4% for the Cohn-Kanade face database and 96.5% for ORL face database. Finally, the face verification application proposed for the Samsung Galaxy SII smart phone indeed successfully verifies a face just in one second which makes it a real-time application.

This paper provides a fingerspelling recognition system with high accuracy rate based on RGB-D image. The system are separated into three parts, including hand region detection, hand feature extraction, and fingerspelling recognition. For the hand regions detection, the regions of hand and face are first obtained by skin color detection and connect component labeling (CCL), and then the hand, the region-of-interest (ROI), is determined by the feature point extraction based on distance transform. Followed is the hand feature extraction which consists of the hand structure and the hand texture. From the feature points of ROI, the locations of palm and fingertips, palm direction, and finger vectors are formed as the hand structure. In addition to the hand structure, this paper adopts the LBP operator to generate the hand texture to deal with the fingerspelling not recognizable by the hand structure. Finally, the extracted hand features are sent into the fingerspelling recognition system, which is built with several different neural network classifiers. The experimental results show that this system is an effective real-time recognition system whose accuracy is higher than 80% for most of the fingerspelling in American Sign Language (ASL).