An Improved Tiered Head Pose Estimation Network with Self-Adjust Loss Function

As an important task in computer vision, head pose estimation has been widely applied in both academia and industry. However, there remains two challenges in the field of head pose estimation: (1) even given the same task (e.g., tiredness detection), the existing algorithms usually consider the estimation of the three angles (i.e., roll, yaw, and pitch) as separate facets, which disregard their interplay as well as differences and thus share the same parameters for all layers; and (2) the discontinuity in angle estimation definitely reduces the accuracy. To solve these two problems, a THESL-Net (tiered head pose estimation with self-adjust loss network) model is proposed in this study. Specifically, first, an idea of stepped estimation using distinct network layers is proposed, gaining a greater freedom during angle estimation. Furthermore, the reasons for the discontinuity in angle estimation are revealed, including not only labeling the dataset with quaternions or Euler angles, but also the loss function that simply adds the classification and regression losses. Subsequently, a self-adjustment constraint on the loss function is applied, making the angle estimation more consistent. Finally, to examine the influence of different angle ranges on the proposed model, experiments are conducted on three popular public benchmark datasets, BIWI, AFLW2000, and UPNA, demonstrating that the proposed model outperforms the state-of-the-art approaches.     

射频前端强电磁脉冲防护模块设计

基于PIN二极管射频限幅的原理,分析了尖峰泄露产生的原因并提出了解决方法。在此基础上采用无源多级PIN二极管结构,提取相应防护电路的S参数,以设计优化匹配网络,研制了工作于1~200 MHz、插入损耗小于0.156 dB、响应时间小于1 ns的射频前端电磁脉冲防护模块。结合PIN二极管模型,利用ADS仿真软件对电磁脉冲防护模块限幅性能进行仿真,并对加工出来的电磁脉冲防护模块进行了测试,结果验证了各项指标满足要求。     

A new form of optical transmission system using 1 × 6 optical switch

Optical switches are useful for various optical systems. This paper describes a new form of optical transmission system, using 1 × 6 optical switch with built-in optical multiplexer. The insertion loss of the optical switch was 3.1 dB (including filter loss of 1.4 dB). This switch is useful for a traffic monitoring system, and its transmission characteristics were measured. The measured signal-to-noise ratio is 44 dB when the switch is applied to a 2-km video transmission line.     

一种铁基超导带材交流损耗的测量

近年来铁基超导材料的发现掀起了又一轮新型超导材料研究的热潮,各种配方的铁基超导材料不断被制备出来,转变温度也不断提高,中国科学家在该领域取得了十分突出的成果,制备出了多种铁基超导材料,但对其交流损耗特性研究鲜有报道,而实际应用中交流损耗的特性又十分重要,因此搭建了一套采用探测线圈法的交流损耗测量平台,测试了一种中国科学院电工研究所采用PIT工艺制作的单芯铁基超导带材的交流损耗。该文介绍了采用探测线圈法测量超导材料交流损耗的一种测量平台的搭建与标定,以及采用标定过的测量平台在液氦温度下测量的一种采用PIT工艺制作的单芯铁基超导带材的交流损耗;并与计算得到的铁基带材超导芯的磁滞损耗和包套的涡流损耗进行了比较分析,得出了其交流损耗与背景磁场频率成正比,其涡流损耗可以忽略的结论。     

会聚区的水平偏移特性及其临界频率分析

利用数值模型研究了完整深海声道中会聚区的水平偏移特性,根据折射定律和Lloyd镜效应推导了会聚区发生水平偏移时临界频率的表达式。研究结果表明:当声波频率低于临界频率时,会聚区的主导模态与频率相关,随着声波频率减小会聚区会向靠近声源的方向水平移动,同时传播损失明显增大,当声波频率大于临界频率时,会聚区的主导模态近似与频率无关,会聚区的位置和传播损失大小不会随频率改变而发生明显变化,声源深度不同时,临界频率也不同。通过临界频率可以确定在特定声源深度下,会聚区发生水平偏移时需要满足的频率条件,利用临界频率与声源深度之间的关系,可以被动估计深海浅层目标的深度。     

用于太赫兹波传输的低损耗、高双折射光纤研究

提出了一种同时具有低吸收损耗、高双折射特性的新型太赫兹传输光纤.光纤由层状的聚四氟乙烯和空气层组成.采用有限元法分析了光纤的结构参数对其吸收损耗和双折射的影响.结果表明:这种光纤的双折射可高达0.08;在保证模式的吸收损耗小于20 dB/m的前提下,其双折射仍可大于0.02. 关键词： 太赫兹传输 吸收损耗 双折射 有限元法     

FEM analysis of magnetic flake composites

A composite comprised of layered flake-like magnetic particles embedded in an insulating medium has been proposed as a low permeability, low loss core material. This would be an alternative to “distributed air gap” compressed powder cores that are widely used for inductors in power applications. Since the lowest loss metallic materials are manufactured in the form of very thin sheets, the particles after pulverizing would be in the form of flakes. The effective permeability and average core loss have been computed for model systems of flake composites in a two-dimensional approximation. The core loss is modeled by eddy current dissipation in the low-frequency limit, where the conductor thickness is much less than the skin depth. It is found that useful values of permeability should be obtained for a modest filling fraction of magnetic material, in contrast to the powder cores which require a value close to unity. The core loss will scale as the inverse of filling fraction, with a small additional enhancement due to perpendicular field components. It is thus expected that useful core materials may be attainable without the necessity of large compaction forces.     

Analytical soliton solutions for the general nonlinear Schrödinger equation including linear and nonlinear gain (loss) with varying coefficients

An improved homogeneous balance principle and an F-expansion technique are used to construct analytical solutions to the generalized nonlinear Schrödinger equation with distributed coefficients and linear and nonlinear gain (or loss). For limiting parameters, these periodic wave solutions acquire the form of localized spatial solitons. Such solutions exist under certain conditions, and impose constraints on the functions describing dispersion, nonlinearity, and gain (or loss). We present a few characteristic examples of periodic wave and soliton solutions with physical relevance.     

光纤偏振效应导致脉冲展宽的解析模型

在10Gb／s,尤其是40Gb／s以上高速光纤通信系统中,光纤的偏振特性已成为限制系统传输距离的主要因素之一。光纤的偏振效应主要包括偏振模色散和偏振依赖损耗。而脉冲均方根展宽是判断信号传输性能的一个主要物理量。本文讨论了光纤线路偏振模色散与偏振相关损耗的相互作用及对信号脉宽的影响。给出了线路偏振模色散矢量和偏振相关损耗矢量之间的关系式,并基于严格的数学方法,导出了在光纤偏振模色散和偏振相关损耗共同作用下的信号均方根脉宽变化的解析形式,同时考虑了光纤色散,啁啾等。该模型可用于分析高阶偏振模色散和偏振相关损耗,任意线性光纤通信系统脉冲展宽分析。     

Reduction of surface transport noise by ground roughness

Measured insertion losses due to the ground effects associated with low configurations of loosely stacked household bricks on a car park are reported. A particularly successful design has the form of a two brick high square lattice which is found to offer a similar insertion loss to regularly-spaced parallel wall arrays of the same height but twice the total width. Part of the insertion loss due to the roughness configurations is the result of transfer of incident sound energy to surface waves which can be reduced by introducing wall absorption or material absorption in the form, for example, of shallow gravel layer. Predicted finite length effects have been explored using a Pseudo-Spectral Time Domain Method, which models the complete 3D roughness profile. It is concluded from measurements and predictions that the lattice design has less dependence on azimuthal source-receiver angle than parallel wall configurations. These predictions are supported by measurements of level difference spectra as a function of azimuthal angle. A 2D Boundary Element Method gives predictions that agree well with data for parallel wall arrays up to 16 m long and it is used to investigate the potential insertion loss of longer configurations up to 0.3 m high. It has been found possible also to make predictions of the insertion loss due to infinitely long 3D lattices using the 2D BEM with the lattice represented by the surface impedance derived from fitting short range data with a slit-pore impedance model. The insertion losses of recessed configurations are predicted to be approximately 3 dB less than those of embossed configurations of the same size. Outdoor experiments also show that pathways can be made through such roughness configurations without significantly affecting their insertion loss. It is concluded that artificial roughness configurations could achieve substantial noise reduction along surface transport corridors without breaking line of sight between source and receiver, thereby proving useful alternatives to noise barriers.     

Energy loss to intravalley acoustic modes in nano-dimensional wire structures at low temperatures

The theory of rate of loss of energy of non-equilibrium electrons due to inelastic interaction with the intravalley acoustic phonons in a nano-dimensional semiconductor wire has been developed under the condition of low lattice temperature, when the approximations of the well known traditional theory are not valid. Numerical results are obtained for narrow-channel GaAs-GaAlAs wires structures. On comparison with other available results it is revealed that the finite energy of the intravalley acoustic phonons and, the use of the full form of the phonon distribution without truncation to the equipartition law, produce significant changes in the energy loss characteristics at low temperatures.     

Analytical modeling of loss characteristics of a polymer arrayed waveguide grating multiplexer

Theoretical analysis is performed for the loss characteristics of a polymer arrayed waveguide grating (AWG) multiplexer around the central wavelength of 1.55 μm with the wavelength spacing of 1.6 nm. The total loss of the device includes the diffraction loss in the input and output (I/O) slab waveguides, bent loss caused by the AWG and I/O channels, leakage loss resulted from the high refractive index substrate, and propagation loss due to the absorption and scattering of the materials of the device. The effects of some structural parameters on the loss characteristics are investigated and discussed. The computed results show that when we select the core thickness as 4 μm, core width as 6 μm, pitch of adjacent waveguides as 15.5 μm, diffraction order as 50, the number of the arrayed waveguides as 91, that of the I/O channels as 17, confined layer thickness between the core and the substrate as 10 μm, distance between the focal point and the origin as 5500 μm, and central angle between the central waveguide and the x-axis (i.e. the vertical of the symmetrical line of the device) as 60°, the total loss of the device can be dropped to the range 3.79–7.93 dB.     

基于猫眼腔镜损耗调节的激光横模演示系统

提出一种基于猫眼腔镜损耗调节的新型激光横模演示系统。采用半外腔He-Ne激光器,利用由分立的可进行五维微调的凸透镜和凹面镜组成的猫眼逆向器构成损耗可调激光谐振腔,通过调节猫眼逆向器改变激光谐振腔的损耗比，实现改变激光横模状态的作用。利用CCD进行图像采集，计算机实现图像存储，从而得到20多种结构清晰的高阶横模花样图片。     

加载形式对涡轮叶尖泄漏损失的影响

本文通过改变90%叶高截面叶型的安装角和厚度分布获得了不同的加载形式,并采用数值模拟方法对比分析了加载形式对叶尖泄漏流动和损失的影响。结果表明,叶尖加载形式影响了泄漏流的流量和法向/流向速度差沿轴向的分布,进而影响泄漏流动损失。随着叶尖负荷向前缘移动,叶尖泄漏总流量增大,当泄漏量和动量差沿轴向分布相对均匀,当地峰值减小,叶中附近所占比重增加;在泄漏量与动量差共同作用下,泄漏涡卷起位置向上游移动,但与主流掺混强度减弱,损失减小;采用均匀加载和前加载形式能有效降低泄漏流与主流的动量差,减小泄漏损失,提高涡轮性能。     

CONCEPTUAL DESIGN OF BESSEL BEAM CAVITY FOR FREE-ELECTRON LASER

The conventional cavity for a free-electron laser (FEL) oscillator usually forms an optical beam of Gaussian mode, which undergoes transverse spread along the interaction region. The transverse divergence of an optical beam will induce reduction of the FEL gain from three aspects: degenerating filling factor, causing diffraction loss and limiting the effective interaction distance. Bessel optical beam has been experimentally demonstrated diffraction-free characteristics in its propagation, which provides a possibility of improvement of FEL gain. In this paper, we present a conceptual design of a Bessel beam cavity for the free-electron laser oscillator. This cavity generates nondiffracting optical beam in the wiggler, resulting in improving the filling factor, decreasing the diffraction loss and elongating the effective interaction distance.     

Insulated electrodes for eliminating conductivity in dielectric relaxation experiments

In a variety of samples, conductivity is a limiting factor regarding the resolution of dielectric loss peaks of interest. One approach to eliminating the signal that originates from conductivity is the use of insulating layers on one or both electrodes. For the typical case, it is shown that insulation layers add errors rather than improving the situation.     

Design and fabrication of electro-optic waveguides with self-assembled superlattice films

This paper presents a strategy for fabricating low loss waveguide devices based on high electro-optic (EO) coefficient self-assembled superlattice (SAS) films, a new sort of polymeric films grown on SiO₂ film and coated with other polymeric films to form multi-layer EO waveguides structure without electric field poling. Firstly, the optical propagation loss induced by the absorption of electrodes is simulated and optimized to obtain both the low optical loss and the low drive voltage. Then this paper gives the scanned electron microscopic (SEM) images of the fabricated devices, the simulated and experimental images of the single guided mode, and the measured optical propagation loss of the EO waveguide devices of 1.0 dB/cm. Finally, the very great agreement between the simulated and measured results of propagation loss of devices is observed.     

Calculating transport AC losses in stacks of high temperature superconductor coated conductors with magnetic substrates using FEM

In this paper, the authors investigate the electromagnetic properties of stacks of high temperature superconductor (HTS) coated conductors with a particular focus on calculating the total transport AC loss. The cross-section of superconducting cables and coils is often modeled as a two-dimensional stack of coated conductors, and these stacks can be used to estimate the AC loss of a practical device. This paper uses a symmetric two dimensional (2D) finite element model based on the H formulation, and a detailed investigation into the effects of a magnetic substrate on the transport AC loss of a stack is presented. The number of coated conductors in each stack is varied from 1 to 150, and three types of substrate are compared: non-magnetic weakly magnetic and strongly magnetic. The non-magnetic substrate model is comparable with results from existing models for the limiting cases of a single tape (Norris) and an infinite stack (Clem). The presence of a magnetic substrate increases the total AC loss of the stack, due to an increased localized magnetic flux density, and the stronger the magnetic material, the further the flux penetrates into the stack overall. The AC loss is calculated for certain tapes within the stack, and the differences and similarities between the losses throughout the stack are explained using the magnetic flux penetration and current density distributions in those tapes. The ferromagnetic loss of the substrate itself is found to be negligible in most cases, except for small magnitudes of current. Applying these findings to practical applications, where AC transport current is involved, superconducting coils should be wound where possible using coated conductors with a non-magnetic substrate to reduce the total AC loss in the coil.     

A Monte Carlo simulation of cathodoluminescence generated in ZnS phosphor powders

Abstract

Field emission displays (FEDs) are the possible full colour flat panel displays of the future. Studies on the FEDs include research on the applicability of wide band gap ZnS phosphors used in cathode ray tubes (CRT's) in the FED environment. Surface reactions occur on the ZnS surface during prolonged electron bombardment to form a non-luminescent ZnO layer with a consequent loss in cathodoluminescence (CL) intensity. Using a Monte Carlo algorithm, the electron energy loss within the ZnS layer of varying thickness on a ZnS substrate and determine the energy loss within the ZnS as function of the ZnO thickness. Since the oxide layer is non-luminescent, the energy loss within the ZnS in proportional to the generated CL. The results obtained correlate with experimental measurements of phosphor degradation during electron bombardment in UHV.     

Nonlinear compression of optical solitons

In this paper, we consider nonlinear Schrödinger (NLS) equations, both in the anomalous and normal dispersive regimes, which govern the propagation of a single field in a fiber medium with phase modulation and fibre gain (or loss). The integrability conditions are arrived from linear eigen value problem. The variable transformations which connect the integrable form of modified NLS equations are presented. We succeed in Hirota bilinearzing the equations and on solving, exact bright and dark soliton solutions are obtained. From the results, we show that the soliton is alive, i.e. pulse area can be conserved by the inclusion of gain (or loss) and phase modulation effects.