Search

Polarization measurement has found applications in many fields ranging from quantum science to our daily life. However, traditional polarization measurement systems suffer from large volume and high cost. To keep pace with the continued device miniaturization and system integration, new detection approaches and devices are desirable. The marriage between optical metasurfaces and superposition of optical vortex beams has offered a new idea method for polarization detection. We propose and experimentally demonstrate a single metasurface for polarization measurement based on the superposition of optical vortex beams. The relationship between the three parameters of the polarization state and the interference pattern of vortex beams is deduced theoretically. The major axis, ellipticity and handedness of a polarization state are determined by analyzing the intensity patterns of the superimposed vortex beams. Our method has provided a new approach for the development of compact optical devices for polarization measurement.

In this paper, we propose a novel heuristics function for evaluating and selecting intermediate points in a goal-driven autonomous exploration and mapping system where navigation is performed by a learned neural network. The function calculation takes into consideration the training setting of the deep reinforcement learning-based network and combines it with distance information towards the global goal and the map information. The candidate with the minimum score is selected as the current intermediate goal. This allows the navigation system to be guided towards the global goal in an informed manner. Experiments in simulation and real-world settings show the benefit of the proposed approach over similar heuristic candidate point evaluation methods.

Traditional surface mounter uses the least square method to calculate the deflection angle of electronic components by fitting a straight line according to the center of a single row of pins on the electronic component. This method can only apply to electronic components with regular pin arrangement, such as SOP (Small Out-Line Package). While the pin distribution of electronic components is irregular, it can’t calculate the deflection angle and center coordinate offset of electronic components. In this paper, an algorithm of rectifying irregular electronic components based on the minimum bounding rectangle fitting is presented. First, basing on the Canny algorithm to get the edge of the irregular electronic components, then the convex hull is constructed on the edge of the components. And finally, the deflection angle and center offset of the irregular electronic components are calculated with the minimum bounding rectangle that fitted by the convex hull of the edge. This method can be applied not only to irregular electronic components, but also to regular electronic components. The experimental results show that the detection precision of the deflection angle can reach 0.1°, the detection error of the center offset can reach 28 μm, and the average processing time of irregular component is 38 ms, which can meet the requirements of the mounter vision system for the high speed and accuracy.

With the increasing demand for state-of-the-art technologies, such as wearable devices and the Internet of things (IoT), power integrity has emerged as a major concern for high-speed, low-power interfaces that are used as mobile platforms. By using case-specific design models in a high-speed memory system, only a limited analysis of the effects of parametric variations can be performed in complex design problems, such as adjacent voltage domain coupling at high frequencies. Moreover, a conventional industrial method can be simulated only after completing the design layout; therefore, a number of iterative back-annotation processes are required for signoff; this delays the time to market. In this paper, we propose a power integrity coanalysis methodology for multiple power domains in high-frequency memory systems. Our proposed methodology can analyze the tendencies in power integrity by using parametric methods, such as parameter sweeping and Monte Carlo simulations. Our experiments prove that our proposed methodology can predict similar peak-to-peak ripple voltages that are comparable with the realistic simulations of low-power double data rate four interfaces.

Recognition of anomalous behavior for crowd gathering and dispersal faces challenges such as target scale change, target short-term disappearance, and target occlusion, which can result in low recognition accuracy. We propose an algorithm based on features of spatial location relationships for the recognition of crowd behaviors to solve the problems mentioned above. We first modified the network based on YOLOV3 by removing the 32x downsampling layer and replacing it with a 4x downsampling layer to further improve the detection of small scale targets, and we re-select the anchor boxes via K-means clustering algorithm to make the network more robust in training. We then perform data association by using the Kalman filter and Hungarian algorithm for target matching between frames to continuously track pedestrians and capture the motion trajectory. The trajectories of pedestrians are transformed to calculate the features of spatial position relationships, and ultimately used as a basis for determining whether abnormal behavior has occurred in the crowd. We validate our approach on PETS2009 dataset and it achieves an 89.0% accuracy with a processing speed of 29.7FPS It is of great significance to maintain the safety of people’s lives and reduce the casualties caused by emergencies in public places.

The DC series arc fault has been a potential fire hazard at the photovoltaic system. The sustained DC series arc can support the enough energy to outbreak of fire so that the proper method to detect the arc fault is needed before the fire ignites. On the contrary of the AC system, the DC series arc don’t have a zero crossing point of the voltage and current. It means that the DC series arc detection is proper to do the detection algorithm based on using the not time domain but frequency domain. The frequency characteristic of the DC series arc fault shows that the certain frequency range helping to classify the arc fault condition or normal condition. The Fast Fourier Transform (FFT) is conducted to make the frequency-based data of the DC series arc fault. To draw a clear frequency feature of the arc, in this paper, the several sampling frequency and sampling numbers are implemented as an analog to digital conversion using TMS320F28033 digital signal processor. The circuit simulating a small photovoltaic system of 300 V 3 A is used to make the actual DC series arc fault and the FFT results of the actual DC series arc are compared and analyzed.

Steady-State Visual Evoked Potentials (SSVEPs) have become one of the most used neural signals for brain- computer interfaces (BCIs) due to their stability and high signal- to-noise rate. However, the performance of SSVEP-based BCIs would degrade with a few training samples. This study was proposed to enhance the detection of SSVEP by combining the supervised learning information from training samples and the unsupervised learning information from the trial to be tested. A new method, i.e. cyclic shift trials (CST), was proposed to generate new calibration samples from the test data, which were furtherly used to create the templates and spatial filters of task- related component analysis (TRCA). The test-trial templates and spatial filters were combined with training-sample templates and spatial filters to recognize SSVEP. The proposed algorithm was tested on a benchmark dataset. As a result, it reached significantly higher classification accuracy than traditional TRCA when only two training samples were used. Speciflcally, the accuracy was improved by 9.5% for 0.7s data. Therefore, this study demonstrates CST is effective to improve the performance of SSVEP-BCI.

In this paper, we present a supervised learning-based mixed-input sensor fusion neural network for autonomous navigation on a designed track referred to as Fusion Drive. The proposed method combines RGB image and LiDAR laser sensor data for guided navigation along the track and avoidance of learned as well as previously unobserved obstacles for a low-cost embedded navigation system. The proposed network combines separate CNN-based sensor processing into a fully combined network that learns throttle and steering angle labels end-to-end. The proposed network outputs navigational commands with similar learned behavior from the human demonstrations. Performed experiments with validation data-set and in real environment exhibit desired behavior. Recorded performance shows improvement over similar approaches.

This paper proposes an effective methodology to figure out the domain of design parameters for given electrical target specifications to have design rules for high-speed package design. Since the exact domain of the parameters is arbitrary and it is difficult to describe accurately, we change this problem by utilizing probability domain. Not only uniformly parametric sampling, but also Monte-Carlo sampling are simulated, and analyzed statistically on probability domain. An optimization algorithm maximizes the parameter domain for a target specification. LPDDR memory channel model constructed for parametric simulation include I/O interface and passive components. Computational script generator for HSPICE simulation is developed to perform parametric simulation and statistical analysis. During the procedure, the problem of using harsh computational burden due to a large number of samplings is resolved by reusing previously sampled data. The proposed methodology can provide a practical and quantitative guideline for high-speed package design. Moreover, the proposed modeling and analysis is unique methodology which is not tied to a specific topology such as conventional modeling and analysis.

We demonstrate a record-high performance monolithic trantenna (transistor-antenna) using 65-nm CMOS foundry in the field of a plasmonic terahertz (THz) detector. By applying ultimate structural asymmetry between source and drain on a ring FET with source diameter $(d_{\text{S}})$ scaling from 30 to $0.37\mu \text{m}$ , we obtained 180 times more enhanced photoresponse $(\Delta u)$ in on-chip THz measurement. Through free-space THz imaging experiments, the conductive drain region of ring FET itself showed a frequency sensitivity with resonance frequency at 0.12 THz in $0.09\sim 0.2$ THz range and polarization-independent imaging results as an isotropic circular antenna. Highly-scalable and feeding line-free monolithic trantenna enables a high-performance THz detector with responsivity of 8.8 kV/ $W$ and NEP of $3.36\text{pW}/\text{Hz}^{05}$ at the target frequency.

This paper shows high-frequency measurement techniques including phase windings and shaft-ends of a three-phase electrical machine rotating at its rated-rpm, and mixed-mode conversion method for impedance parameters. For the purpose, an external torque source mechanically drives the induction machine, while measuring phase windings and shaft-ends with a vector network analyzer. To investigate common-mode, differential-mode, and inter-phase differential-mode components of drive signals with each shaft-end, we converted the measured impedance parameters to mixed-mode impedance parameters. To verify the converted results, transfer impedances between shaft ends and common-mode component of the mixed-mode parameters are compared with time-domain shaft voltage waveforms that are generated by sinusoidal CM input signals of different frequencies. The comparison showed the mixed-mode impedance parameters match the time-domain waveforms. Also, the mixed-mode parameters showed that common- and differential-mode are more responsible for shaft end-to-end voltage, while common-mode and inter-phase differential-mode are more responsible for shaft-to-frame voltage, for the induction machine. Also, frequency ranges of each mode over which each voltage are more susceptible are specified from the results.

The power integrity of high-speed interfaces is an increasingly important issue in mobile memory systems. However, because of complicated design variations such as adjacent VDD domain coupling, conventional case-specific modeling is limited in analyzing trends in results from parametric variations. Moreover, conventional industrial methods can be simulated only after the design layout is completed and it requires a lot of back-annotation processes, which result in delayed delays time to market. In this paper, we propose a chip-package-PCB coanalysis methodology applied to our multi-domain high-speed memory system model with a current generation method. Our proposed parametric simulation model can analyze the tendency of power integrity results from variable sweeps and Monte Carlo simulations, and it shows a significantly reduced runtime compared to the conventional EDA methodology under JEDEC LPPDR4 environment.

Malware remains a serious problem for economic corporations, government agencies, and individuals. With the evolution of malware, the malware detection method based on signature, reputation or anomaly show the lack of detection ability. In this paper, we proposed a novel malware detection system, which use the feature extract from intermediate representation and utilizes multiple evaluation methods to detect the malware. We extract the feature from intermediate representation with fine-grained that retain the functional sequence of malware. Then, we use LSTM (Long Sort Term Memory) to learn the pattern from the training set and build a detection system that contain many antivirus software's results to detect malware and classify it. Our approach is evaluated on more than 300,000 samples. Experimental results show that our system can effectively detect unknown malware and classify it.

The coupling of power domains in a typical package power distribution network (PDN) is investigated, mainly focusing on the effect of ground structures and substrate parameters. From the simulated impedance parameter data, the self and the coupling capacitances are calculated. The preliminary study shows that the effect of the bottom ground planes has the most dominant effect on the domain coupling, and the ground plane between adjacent power domains can be effectively used to reduce the domain coupling.

We report a highly-sensitive plasmonic nano-ring transistor for monolithic terahertz (THz) active antenna. By designing an ultimate asymmetric transistor on a metal-gate structure, more enhanced (180 times) channel charge asymmetry has been obtained in comparison with a bar-type asymmetric transistor of our previous work. In addition, by exploiting ring-type transistor itself as a monolithic circular active antenna, which is designed for a 0.12-THz resonance frequency, we experimentally demonstrated the highly-enhanced responsivity (R V ) > 1 kV/W (x 5) and reduced noise-equivalent power (NEP) < 10 pW/Hz0.5 (x 1/10).

We report a circular-shape monolithic transistor-antenna (trantenna) for high-performance plasmonic terahertz (THz) detector. By designing an asymmetric transistor on a ring-type metal-gate structure, more enhanced (45 times) channel charge asymmetry has been obtained in comparison with a bar-type asymmetric transistor of our previous work. In addition, by exploiting ring-type transistor itself as a monolithic circular patch antenna, which is designed for a 0.12-THz resonance frequency, we demonstrated the highly-enhanced responsivity (R V ) > 1 kV/W (× 5) and reduced noise-equivalent power (NEP) < 10 pW/Hz0.5 (× 1/10).

Visual brain-computer interfaces (BCIs) have achieved great progress in speed recently. But the problem of visual fatigue caused by intense flashes poses a great challenge in designing practical systems for long-term use. A direct way to improve visual comfort is to reduce the stimulus contrast. But it could also weaken the featured evoked potentials, which would bring a negative impact on system accuracy. Thus it's significant to figure out the optimal contrast that could have both high visual comfort and high accuracy. This study investigated the effects of different stimulus contrasts on the two aspects. Six hybrid P300-SSVEP spellers were developed with different stimulus contrasts. Three subjects spelled 10 same characters offline for each speller. After each spelling subjects were asked to grade the flashes they just met in terms of visual comfort. Stepwise linear discriminant analysis (SWLDA) was used to recognize the P300 potential; the filter bank canonical correlation analysis (FBCCA) with individual template was adopted to classify the SSVEP. A decision fusion was performed to recognize the target. The results showed that, compared with P300 or SSVEP only features, the hybrid features significantly improved the accuracy. The subjects felt more comfortable for contrasts below 25%. The classification accuracy wouldn't have a great loss unless the contrast was below 12%. Taken together, the optimal contrast was 12% for the hybrid P300-SSVEP BCI system in this study.

In this paper an algorithm is described for obtaining the response of Power Distribution Networks (PDN) arising in chip, package or pcb during an early design phase. Results are provided for power grids arising in silicon interposers to validate the approach.

For the electrical modeling of electronic packaging composed of planar and vertical interconnections, this paper presents the combination of two integral equation based circuit extraction methods. Each method, which geometrically accords with planar or cylindrical structures, can generate equivalent circuit elements separately. The planar and cylindrical interconnection models are connected in circuit level, and the coupling between the two models can be considered by additional capacitive components.

We propose a topology-based Bayesian visual navigation framework, with which we represent the world environment as a collection of scenes. The proposed topological map consists of nodes represented as a bag of visual line words and edges represented as both an adjacency list and relative motion information to perform the transition between topological nodes. Our proposed Bayesian localization framework uses two measurement models: a visual line word-based place model and a path-matching model. To enable a mobile robot to reach to a desired destination, a coastal path is planned in such a way that maximizes the possibility of following the reference path defined as a sequence of scenes encountered when building the topological map, while avoiding possible collisions by leveraging a local grid map constructed in real time from Kinect depth information. To show the validity of our proposed framework, we provide several extensive experimental results in several indoor environments under different conditions, such as illumination changes and visual occlusions.

Compared with traditional cryptography, biometric cryptosystems provide more convenient and safe protection of keys. In this paper, an effective scheme for biometric cryptosystems based on the intersections of hyper-plane and hashed discrete space is proposed. Only genuine feature vectors, which are located on the hyper-plane, find the correct solution. Several tricks are provided to ensure the security of proposed scheme. Experiments have shown that our method achieved approximate performance compared with one of the latest methods based on fuzzy vault. However, our method is less complex than fuzzy vault based methods.

In this paper, we propose a scene clustering algorithm which uses straight line features. Scenes are represented as nodes in the graph, and each connectivity between nodes is calculated by a pre-trained vocabulary tree. By applying a spectral clustering algorithm to the constructed graph, the scenes are partitioned into k groups where k is determined by the proposed estimation method. Instead of using the standard eigenvalue analysis, the optimal k is computed so that the partitioned graph becomes to have strong intra-class correlations while inter-class correlations are relatively weak. As a result, scenes are adaptively clustered according to the environmental changes. The clustering performance of the proposed method is quantitatively evaluated with three image sequences captured in challenging environments. Experimental comparisons demonstrate that our line-based algorithm outperforms existing algorithms utilizing other types of features as well as the produced scene clustering results are more human-like.

Since early 1990, event recognition has been one of the most attractive research topics for video understanding, in company with object recognition. Most studies on video event recognition, which are based on data-driven approaches, should train a model for a newly-added event without using human knowledge and existing models for similar events. Because it is impossible to define all events required for video understanding in advance, this paper proposed a hierarchical recognition method for general events based on dynamic spatial relations between two objects and specialized events determined by the related objects. The general events are useful for describing interactions between objects of interest regardless of video domain. The specialized events can be provided to users as familiar terms in video interpretation or visual question answering for user-friendly interaction. For two general events and their specialized four events, the proposed recognition method performed the F-score of 82.31% and 88.61% based on object-based and region-based event matching, respectively.

A modified square Koch curve fractal antenna design working as a stent for partially or even fully biocompatible wireless implants is presented. The idea of using stent as a antenna, is to enable data transfer and wireless power transfer for charging integrated circuit device battery, to avoid a surgery for changing the battery after its period of use. This study provides a state-of-the-art overview that describes the key technological challenges and a design methodology for loop type modified Koch curve stent antenna. The simulation studies of our 730 MHz stent based antenna, implanted at a depth of 5.5 mm with in hand (Below elbow), showed a 20.58 dB power reduction as compare to free space interface stent antenna design. These results are evaluated by using a high-fidelity simulation model including components of Human Body model to provide the effects of tissues on stent antenna system. In order to validate the idea of using Koch fractal curves as a loop type stent antenna, the frequency domain characteristics of the proposed antenna system are presented.

To predict the radiated emission (RE) from AC motors in driver systems under the CISPR 25 standard, this paper proposes a method to estimate the radiation pattern of the stator winding by using antenna array theory. The availability of applying the proposed method is ensured by identifying the radiation region boundaries of the winding structure of a reference motor. The array model of the stator winding can be constructed by magnetic vector potentials from the circular array of axial and end-winding current elements. The calculated patterns, which are excited by the balanced three-phase harmonics at the resonance frequency, show that the radiation becomes significant when the harmonic number is not a multiple of three. Also, we can find that the radiation from the end-winding currents is dominant. Since the estimated radiation patterns are fairly correlated with 3D full-wave simulation results, the proposed method is expected to be used for the prediction of the RE from any stator winding configuration.

This paper proposes an improved high-frequency equivalent model of ac machine, focusing on the accurate extraction of stator-to-frame capacitance. The behavior that the capacitance decreases in several stages throughout the observed frequency is analyzed and modeled by applying Debye model. To verify the constructed equivalent model, the input and transfer impedances of a phase line is compared with measurement. The newly developed per-unit-length equivalent model with the improved admittance model can be used to estimate the high-frequency behavior of ac motor accurately.

In this paper we provide a method that can be used to analyze power distribution networks with electromagnetic accuracy. The method is applicable to any power distribution network in the chip, package or printed circuit board (PCB). The method can also be used to compute the interaction of the power distribution network between various domains. We provide examples to validate this approach.

To construct design rules for high-speed I/O interfaces, a method to identify the domain of design parameters for given electrical specifications is proposed in this paper. Since the shape of parameter domain is arbitrary and difficult to describe accurately, we simplify the problem by considering cubic domains only. Each domain is tested whether it belongs to the original parameter domain by Monte-Carlo sampling, and the volume of the domain is maximized by an optimization algorithm. During the procedure, computational burden due to a large number of samplings is addressed by reusing previously sampled data. By applying the developed method to a typical memory I/O interface model, we show the proposed approach can provide a practical and quantitative guideline for high-speed channel design.