Rotation,Translation and Scale Invariant Sign Word Recognition Using Deep Learning

Communication between people with disabilities and people who do not understand sign language is a growing social need and can be a tedious task.One of the main functions of sign language is to communicate with each other through hand gestures.Recognition of hand gestures has become an important challenge for the recognition of sign language.There are many existing models that can produce a good accuracy,but if the model test with rotated or translated images,they may face some difficulties to make good performance accuracy.To resolve these challenges of hand gesture recognition,we proposed a Rotation,Translation and Scale-invariant sign word recognition system using a convolu-tional neural network(CNN).We have followed three steps in our work:rotated,translated and scaled(RTS)version dataset generation,gesture segmentation,and sign word classification.Firstly,we have enlarged a benchmark dataset of 20 sign words by making different amounts of Rotation,Translation and Scale of the ori-ginal images to create the RTS version dataset.Then we have applied the gesture segmentation technique.The segmentation consists of three levels,i)Otsu Thresholding with YCbCr,ii)Morphological analysis:dilation through opening morphology and iii)Watershed algorithm.Finally,our designed CNN model has been trained to classify the hand gesture as well as the sign word.Our model has been evaluated using the twenty sign word dataset,five sign word dataset and the RTS version of these datasets.We achieved 99.30%accuracy from the twenty sign word dataset evaluation,99.10%accuracy from the RTS version of the twenty sign word evolution,100%accuracy from thefive sign word dataset evaluation,and 98.00%accuracy from the RTS versionfive sign word dataset evolution.Furthermore,the influence of our model exists in competitive results with state-of-the-art methods in sign word recognition.

Upper bound analysis for deep tunnel face with joined failure mechanism of translation and rotation

A joined failure mechanism of translation and rotation was proposed for the stability analysis of deep tunnel face, and the upper bound solution of supporting force of deep tunnel was calculated under pore water pressure. The calculations were based on limit analysis method of upper bound theory, with the employment of non-associated Mohr-Coulomb flow rule. Nonlinear failure criterion was adopted. Optimized analysis was conducted for the effects of the tunnel depth, pore water pressure coefficient, the initial cohesive force and nonlinear coefficient on supporting force. The upper bound solutions are obtained by optimum method. Results are listed and compared with the previously published solutions for the verification of correctness and effectiveness. The failure shapes are presented, and results are discussed for different pore water pressure coefficients and nonlinear coefficients of tunnel face.

Rotational and translational domains of beta precipitate in aged binary Mg-Ce alloys

The structural evolution fromβ_(1)(Mg_(3)Ce)toβ(Mg_(12)Ce)precipitates,which takes place at the over-aged stage of binary Mg-Ce alloys,are investigated by high-angle annular dark-field scanning transmission electron microscopy.The structural transformation mainly occurs in the{111}_(β1)crystallographic planes,where the newly formedβlattices exhibit two categories of domain structures,namely rotational and translational domains.The rotational domain is composed of threeβdomains(β_(RA),β_(RB)andβ_(RC)),which are related by a 120°rotation with respect to each other around the 111_(β1)axis of theirβ_(1)parent phase.The{111}_(β1)crystallographic planes can provide four sets of sublattices with the same orientation for an initial nucleation ofβlattice.It leads to the formation of four translationalβdomains(β_(TA),β_(TB),β_(TC)andβ_(TD)),among which any two differ by a vector of 1/6112_(β1).We deduce theoretically that there exist twenty-fourβdomains during this transition.However,considering the interfacial misfit,only one-third of domains can grow up and eventually formsβribbon.Furthermore,a majority ofβribbons overlap partiallyβ_(1)plate,which is beneficial to relax interfacial strain amongβ,β_(1)andα-Mg matrix(α/β/β_(1)).The configuration of multipleβdomains can effectively regulate interfacial misfit ofα/βandβ/β_(1),which are responsible for enhancing the hardness and strength of Mg-Ce alloy.Additionally,this study aims to provide some clues to improve the over-aged performance of magnesium alloys by constructingβdomains and optimizing theα/β/β_(1)interface.

Image Reconstruction from Fan-Beam Projections without Back-Projection Weight in a 2-D Dynamic CT: Compensation of Time-Dependent Rotational, Uniform Scaling and Translational Deformations

In a dynamic CT, the acquired projections are corrupted due to strong dynamic nature of the object, for example: lungs, heart etc. In this paper, we present fan-beam reconstruction algorithm without position-dependent backprojection weight which compensates for the time-dependent translational, uniform scaling and rotational deformations occurring in the object of interest during the data acquisition process. We shall also compare the computational cost of the proposed reconstruction algorithm with the existing one which has position-dependent weight. To accomplish the objective listed above, we first formulate admissibility conditions on deformations that is required to exactly reconstruct the object from acquired sequential deformed projections and then derive the reconstruction algorithm to compensate the above listed deformations satisfying the admissibility conditions. For this, 2-D time-dependent deformation model is incorporated in the fan-beam FBP reconstruction algorithm with no backprojection weight, assuming the motion parameters being known. Finally the proposed reconstruction algorithm is evaluated with the motion corrupted projection data simulated on the computer.

Translation Elongation by Rotation in a Still Fluid of a Solid Cone with a Spiral Vane

Within the field of physics, the idea of perpetual motion is not allowed in most references. However, in a still fluid, one potential perpetual motion example is offered: a hollow solid cone rotating about its long axis at a constant rate.

NCCT for Micropolar Solid and Fluid Media Based on Internal Rotations and Rotation Rates with Rotational Inertial Physics: Model Problem Studies

This paper presents model problem studies for micropolar thermoviscoelastic solids without memory and micropolar thermoviscous fluid using micropolar non-classical continuum theories (NCCT) based on internal rotations and rotation rates in which rotational inertial physics is considered in the derivation of the conservation and balance laws (CBL). The dissipation mechanism is due to strain rates as well as rotation rates. Model problems are designed to demonstrate and illustrate various significant aspects of the micropolar NCCT with rotational inertial physics considered in this paper. In case of micropolar solids, the translational and rotational waves are shown to coexist. In the absence of microconstituents (classical continuum theory, CCT) the internal rotations are a free field, hence have no influence on CCT. Absence of gradients of displacements and strains in micropolar thermoviscous fluid medium prohibits existence of translational waves as well as rotational waves even though the appearance of the mathematical model is analogous to the solids, but in terms of strain rates. It is shown that in case of micropolar thermoviscous fluids the BAM behaves more like time dependent diffusion equation i.e., like heat conduction equation in Lagrangian description. The influence of rotational inertial physics is demonstrated using BLM as well as BAM in the model problem studies.

Multi-Barycenter Mechanics, N-Body Problem and Rotation of Galaxies and Stars

In the present paper, the establishment of a systematic multi-barycenter mechanics is based on the multi-particle mechanics. The new theory perfects the basic theoretical system of classical mechanics, which finds the law of mutual interaction between particle groups, reveals the limitations of Newton’s third law, discovers the principle of the intrinsic relationship between gravity and tidal force, reasonably interprets the origin and change laws for the rotation angular momentum of galaxies and stars and so on. By applying new theory, the multi-body problem can be transformed into a special two-body problem and for which an approximate solution method is proposed, the motion law of each particle can be roughly obtained.

Hierarchical path planning for multi-arm spacecraft with general translational and rotational locomotion mode

On-orbit construction and maintenance technology will play a significant role in future space exploration.The dexterous multifunctional spacecraft equipped with multi-arm,for instance,Spider Fab Bot,has attracted a great deal of focus due to its versatility in completing these missions.In such engineering practice,point-to-point moving in a complex environment is the fundamental issue.This paper investigates the three-dimensional point-to-point path planning problem,and a hierarchical path planning architecture is developed to give the trajectory of the multi-arm spacecraft effectively and efficiently.In the proposed 3-level architecture,the high-level planner generates the global constrained centric trajectory of the spacecraft with a rigid envelop assumption;the middle-level planner contributes the action sequence,a combination of the newly developed general translational and rotational locomotion mode,to cope with the relative position and attitude of the arms about the centroid of the spacecraft;the low-level planner maps the position/attitude of the end-effector of each arm from the operational space to the joint space optimally.Finally,the simulation experiment is carried out,and the results verify the effectiveness of the proposed three-layer architecture path planning strategy.

Translation and rotation of particles in different flow pattern areas of a silo

The present paper reports the results obtained for translational and rotational velocity profiles of spherical particles for the mixed flow in a conical silo.The discrete element method(DEM)based on Hertz-Mindlin(no slip)with RVD rolling friction contact model is used for simulations.Opposite correlations are found between translational and rotational velocities in different flow areas of the silo.In particular,the abrasion caused by rotation is dominant in the funnel flow area.In addition,increase of the mass flow rate of silo can effectively reduce the abrasion induced by rotation.This highlights that understanding of dynamic characteristics of particles is helpful for optimization of silos and reduction of granular material abrasion.

A Cone Rotating in a Fluid Translates

When a solid cone with smooth surfaces rotates at a constant rate about its long axis in still water, it will experience no friction, as announced earlier [1] and documented further here. However, if not restrained, it will translate along the axis base first and apex last, which is caused by a variation of pressure on its side: low near the base and high at the apex. This translation needs to be verified experimentally. Friction will occur during translation but an opposite reaction force will also take place related to the front to back asymmetry of the body [2]. Whether or not these two oppositely directed forces can cancel each other out by an appropriate choice of variable magnitudes is not known, but if true, it would lead to an extended translation. Also the translation path could be lengthened by slightly modifying the front face of the cone such that the translation itself produces rotation. One idea for such a modification is suggested. Observations in the future would be very beneficial.

Rotating Translating Cones

A first attempt has been made to confirm experimentally a theoretical concept, recently published, involving a rigid cone rotating about its long axis under still water: it should tend to translate along that axis blunt end leading and apex trailing. Two identical hollow cones, neutrally buoyant, with equal weights attached to the apexes, were released simultaneously at the surface of a swimming pool. One cone had a thin light weight spiral vane vertically attached to the cone’s outside surface in order to cause it to rotate as it sank. Several trial runs were made in the shallow and deep ends of the pool, and in every case, the non-rotating cone without a vane hit the bottom of the pool first. These comparisons qualitatively and indirectly validate the prediction.

基于三维空间旋转平移的自适应知识表示方法

现有知识图谱表示学习研究中普遍存在忽视特定关系的语义空间、难以建模非单射复杂关系或多种关系模式等问题,尤其是在不可交换的组合以及子关系两种关系模式上表现不佳。针对该问题,在对实体自适应投影的基础上,利用罗德里格斯旋转公式将旋转操作从二维空间拓展到三维空间并进行平移优化,提出一种新的具有强表征能力的模型ATR3DKRL。通过理论推导可以证明该模型能够建模非单射复杂关系以及多种关系模式。在多个通用数据集上的实验结果表明,该模型可以有效提高链接预测精度,在数据集DB100K与FB15K-237中四个指标上领先现有基线模型,其中在DB100K上评价指标MRR和H@1相较于基线模型RotatE分别大幅提高了3.3%以及6.5%。

一类第二型曲线积分的四种计算方法

针对一类第二型曲线积分的计算问题,本文给出了四种计算方法.总体上这四种解法可归纳为三类:第一类是转化为第一型曲线积分,第二类是积分路线参数化,第三类是化曲线积分为曲面积分的Stokes方法.通过这些解法的讨论,有利于启发和提高学生综合分析问题的能力.

三枕捣固作业过程中道砟细观运动及能量演变分析

使用大型捣固车对有砟轨道进行作业,是改善道床质量状态最有效方法和通用手段,是保障列车安全运营的必然选择。但我国大型养路机械源于引进国外技术,对捣固车捣固作业机理认识不足,无法进行科学养护维修。为加深对目前铁路养修作业中常用DWL-48捣固稳定车的三枕捣固装置作业过程的认识,该文借助离散元与多体动力学协同仿真分析方法,建立了三枕捣固装置-轨排-有砟道床三维空间精细化耦合仿真模型,并结合现场实测结果验证了模型的正确性,分析了捣固作业动态过程中道砟颗粒平动、转动特性及能量演变规律。研究结果表明:捣镐对道砟颗粒平动、转动和能量的影响主要位于枕下0 mm~175 mm区域;道砟颗粒速度与角速度变化规律具有很好的同步性,有助于快速填充枕下空隙。捣固作业过程中捣镐采用“道砟平动为主,转动为辅”的方式填充枕下空隙,并且在该过程中伴随着道砟颗粒平动动能、旋转动能逐渐向道砟颗粒势能的不断演变。

刚毛翼拍合过程中复合运动对升力的影响

为揭示拍合过程中拍动和俯仰运动重叠对刚毛翼升力的影响,基于重叠网格方法建立了刚毛翼的计算流体动力学模型,分析了刚毛翼悬飞时的黏性效应,在此基础上研究了拍动-俯仰重叠率对刚毛翼升力系数的影响规律及机制。研究结果表明:随着基于弦长的雷诺数由10逐渐增加至80,刚毛间隙泄漏明显增强,刚毛翼内外侧压差和刚毛间隙泄漏率与雷诺数之比不断减小,导致压力和剪切力对升力系数的贡献不断减小,升力系数不断降低;通过采用较高的拍动-俯仰重叠率,可使拍合过程中的流体泄漏方向与刚毛运动方向相反,在合拢阶段形成明显的竖直向上剪切力合力,而在打开阶段形成明显的竖直向上压差,从而将合拢和打开阶段的负升力峰值转变为正升力峰值,有效提高刚毛翼的平均升力系数。

基于二维正态分布的无人机激光雷达点云匹配研究

无人机激光雷达点云特征多,一次匹配花时间长,难以进行二次匹配,因此研究基于二维正态分布的无人机激光雷达点云匹配方法。采集无人机激光雷达点云图像,通过旋转平移方法、双边滤波方法对图像预处理,利用二维正态分布算法和动态时间规整算法完成点云特征提取,使用初始变换矩阵估计算法对点云进行粗匹配,再使用近点迭代算法进行点云快速精匹配,通过两次匹配实现无人机激光雷达点云快速匹配。实验结果表明,所提方法的无人机激光雷达点云图像去噪效果好,点云匹配时间短,匹配偏差仅在0.04 m-0.15 m之间,匹配精度达到了相关预期。

基于运动形式转化利用转动惯量虚拟平动惯性质量的TMD控制系统

针对调谐质量阻尼器(Tuned Mass Damper,TMD)通常需要较大的附加质量,安装空间受限以及质量块运动时需要较大的行程等问题,基于平动‐转动运动形式相互转化和能量守恒原理,本文提出了利用转动惯量虚拟平动惯性质量的TMD控制系统(Rotary inertia virtualizing translational mass based Tuned Mass Damper,简称RTMD),进行RTMD控制系统的设计概念,以单自由度结构对象为例建立了附加RTMD控制系统的运动方程,分析了RTMD控制系统参数对结构振动控制效果的影响规律。结果表明控制效果与系统的质量比、惯质比、阻尼比等参数密切相关,相关规律也可以推广到一般多自由度结构体系一阶振动的调谐吸振减振控制。进行了模型振动台试验研究,时域分析和频域分析结果均表明,试验结果与基于理论模型的数值分析结果一致性良好,验证了RTMD控制系统理论模型的正确性、设计参数的合理性以及控制系统应用于实际问题的可行性。

不同体积脑转移瘤对容积旋转调强计划摆位误差的敏感性

目的:探讨不同体积脑转移瘤对容积旋转调强(VMAT)计划摆位误差的敏感性。方法:在头部CT图像上设置5个球形模拟多发脑转移瘤,体积依次为0.5 cm^(3)(V1)、1.7 cm^(3)(V2)、4.0 cm^(3)(V3)、8.1 cm^(3)(V4)和14.0 cm^(3)(V5)。制定单等中心5弧非共面VMAT计划,等中心点到各靶区的距离为5 cm。通过旋转治疗床模拟绕Z轴旋转的摆位误差θ(-3.0°~3.0°,步长为0.5°),通过移动等中心在三维轴向(X/Y/Z)上的位置模拟平移误差(-3.0 mm~3.0 mm,步长为0.5 mm),不改变优化条件的情况下,重新计算剂量分布。记录旋转、平移后靶区的覆盖率,并归一到未发生摆位误差下的相对覆盖率。运用线性回归法,分析不同体积脑转移瘤靶区覆盖率与体积的相关性。结果:随着旋转误差的增加,靶区相对覆盖率逐渐下降。恒定旋转误差情况下,靶区相对覆盖率与体积呈线性相关(θ=0.5°,P=0.006;θ=1.0°,P=0.024;θ=1.5°,P=0.028;θ=2.0°,P=0.019;θ=2.5°,P=0.014;θ=3.0°,P=0.007),随着靶区体积增加,靶区相对覆盖率逐渐上升。当旋转误差θ<0.5°时,无论大体积靶区(14.0 cm^(3))还是小体积靶区(0.5 cm^(3)),其靶区相对覆盖率变化均小于2%。随着平移误差的增大,靶区相对覆盖率逐渐下降。相对于大体积靶区,小体积靶区对平移误差更为敏感。当平移误差<0.5 mm时,大体积靶区相对覆盖率变化<2%,而小体积靶区(0.5 cm^(3))的相对覆盖率下降接近5%。结论:旋转误差恒定情况下,靶区相对覆盖率与体积呈线性关系。相对于大体积靶区,小体积靶区对平移误差更为敏感。建议对于小体积靶区在实施立体定向放射外科VMAT计划时,旋转误差应控制在0.5°以内,平移误差应控制在0.5 mm以内。

Scoliocorrector Fatma-UI for correction of adolescent idiopathic scoliosis: Development, effectivity, safety and functional outcome

BACKGROUND Adolescent idiopathic scoliosis remains a major problem due to its high incidence,high risk,and high cost.One of the aims of the management in scoliosis is to correct the deformity.Many techniques are available to correct scoliosis deformity;however,they are all far from ideal to achieve three-dimensional correction in scoliosis.AIM To develop a set of tools named Scoliocorrector Fatma-UI(SCFUI)to aid threedimensional correction and to evaluate the efficacy,safety,and functional outcome.METHODS This study consists of two stages.In the first stage,we developed the SCFUI and tested it in finite element and biomechanical tests.The second stage was a single-blinded randomized clinical trial to evaluate the SCFUI compared to direct vertebral rotation(DVR).Forty-four subjects with adolescent idiopathic scoliosis were randomly allocated into the DVR group(n=23)and SCFUI group(n=21).Radiological,neurological,and functional outcome was compared between the groups.RESULTS Finite element revealed the maximum stress of the SCFUI components to be between 31.2-252 MPa.Biomechanical analysis revealed the modulus elasticity of SCFUI was 9561324±633277 MPa.Both groups showed improvement in Cobb angle and sagittal profile,however the rotation angle was lower in the SCFUI group(11.59±7.46 vs 18.23±6.39,P=0.001).Neurological and functional outcome were comparable in both groups.CONCLUSION We concluded that SCFUI developed in this study resulted in similar coronal and sagittal but better rotational correction compared to DVR.The safety and functional outcomes were also similar to DVR.

系列小分子液体中α弛豫与探针离子电导行为的对比

液体中平动和转动的耦合性是凝聚态物理长期关注的问题之一,本文采用介电谱方法同时获得了系列小分子液体中α弛豫的弛豫时间和探针离子的电导率.样品包括具有不同分子形状和官能团的碳原子数跨度在3—14范围内的15种一元和二元小分子液体.分析结果表明平动和转动的耦合性与液体分子的官能团并没有直接的对应关系,对分子形状、大小和离子大小也不是十分敏感,但是液体的微观结构是影响平动和转动耦合性的重要因素.也就是,无论在一元还是二元体系中,液体的微观结构没有改变时电导率的倒数和弛豫时间与温度的依赖关系具有一致性,这也为弛豫时间的测量提供了一种方法.研究结果还表明,液体中自身携带的杂质离子与定量掺入离子的电导率的温度依赖关系相同,为电解质溶解度低的有机小分子液体中离子电导率行为的研究提供了思路.本文中单羟基醇的实验结果也与单羟基醇中α弛豫而非Debye弛豫对应于体系结构弛豫的观点相一致.