基于Point Transformer方法的鱼类三维点云模型分类

为实现对不同鱼类的精准分类,研究共采集110尾真实鱼类的三维模型,对获取的3D模型进行基于预处理、旋转增强和下采样等操作后,获取了1650尾实验样本。然后基于Point Transformer网络和2个三维分类的对比网络进行数据集的分类训练和验证。结果表明,利用本实验的目标方法Point Transformer获得了比2个对比网络更好的分类结果,整体的分类准确率能够达到91.9%。同时对所使用的三维分类网络进行有效性评估,3个模型对于5种真实鱼类模型的分类是有意义的,其中Point Transformer的模型ROC曲线准确率最高,AUC面积最大,对于三维鱼类数据集的分类最为有效。研究提供了一种可以实现对鱼类三维模型进行精准分类的方法,为以后的智能化渔业资源监测提供一种新的技术手段。

Optimizing electronic structure through point defect engineering for enhanced electrocatalytic energy conversion

Point defect engineering endows catalysts with novel physical and chemical properties,elevating their electrocatalytic efficiency.The introduction of defects emerges as a promising strategy,effectively modifying the electronic structure of active sites.This optimization influences the adsorption energy of intermediates,thereby mitigating reaction energy barriers,altering paths,enhancing selectivity,and ultimately improving the catalytic efficiency of electrocatalysts.To elucidate the impact of defects on the electrocatalytic process,we comprehensively outline the roles of various point defects,their synthetic methodologies,and characterization techniques.Importantly,we consolidate insights into the relationship between point defects and catalytic activity for hydrogen/oxygen evolution and CO_(2)/O_(2)/N_(2) reduction reactions by integrating mechanisms from diverse reactions.This underscores the pivotal role of point defects in enhancing catalytic performance.At last,the principal challenges and prospects associated with point defects in current electrocatalysts are proposed,emphasizing their role in advancing the efficiency of electrochemical energy storage and conversion materials.

Sensitivity Analysis of Structural Dynamic Behavior Based on the Sparse Polynomial Chaos Expansion and Material Point Method

This paper presents a framework for constructing surrogate models for sensitivity analysis of structural dynamics behavior.Physical models involving deformation,such as collisions,vibrations,and penetration,are devel-oped using the material point method.To reduce the computational cost of Monte Carlo simulations,response surface models are created as surrogate models for the material point system to approximate its dynamic behavior.An adaptive randomized greedy algorithm is employed to construct a sparse polynomial chaos expansion model with a fixed order,effectively balancing the accuracy and computational efficiency of the surrogate model.Based on the sparse polynomial chaos expansion,sensitivity analysis is conducted using the global finite difference and Sobol methods.Several examples of structural dynamics are provided to demonstrate the effectiveness of the proposed method in addressing structural dynamics problems.

基于注意力权重PointNet++的电力走廊点云语义分割研究

传统的电力走廊点云数据的分割会出现精度低、数据局部特征捕获存在局限性等问题,为此提出了一种基于注意力权重的PointNet++网络场景分割模型。将深度学习中的PointNet++算法用于电力走廊场景分割中,再引入了空间注意力机制,帮助模型更有效地关注重要的空间区域。为此采用自制的数据集,并基于PointNet++网络模型的经典结构,在每个点集抽取(set abstraction,SA)模块中的多层感知机(multi layer perceptron,MLP)加入倒置瓶颈设计,提高对点云数据的处理效率和准确性。研究结果表明,与传统的PointNet++网络相比,改进的PointNet++网络平均交并比(mean intersection over union,mIoU)高出6.3%,加入空间注意力机制的改进模型在自制数据集上表现出更好的分割效果,尤其是在边界划分方面提升明显,验证了该方法在点云语义分割上的有效性。

位置自适应卷积PointNet++的点云数据分类方法

针对复杂场景中点云数据分类精度低问题,提出一种基于位置自适应卷积的PointNet++深度神经网络模型。由于位置自适应卷积具有较强捕捉细粒度局部特征能力,能充分获取三维点云的空间变化和几何结构特征信息,故本文在PointNet++基础上,首先通过最远点采样获取关键点,其次根据关键点使用K最近邻方法(KNN)实现分组,然后由位置自适应卷积代替原方法中的MLP提取每组的局部特征,最终完成点云分类。在2个公开的点云数据集S3DIS、Semantic3D上对本文方法进行多次对比实验,实验结果表明,本文方法在室内数据集S3DIS上的总体精度和mIoU较PointNet++网络分别提高约2.7个百分点和3.2个百分点,在室外数据集Semantic3D上的总体精度和mIoU PointNet++分别高出约2.5个百分点和2.1个百分点。

Effect of acupuncture on acupoint "Yingxiang-Hegu" on Th1, Th2 cytokines and T-bet/GATA-3 of allergic rhinitis rats

Objective:To explore the effect of"Yingxiang-Hegu"on Th1,Th2-related cytokines and[2]transcription factors T-bet and GATA-3 in rats with allergic rhinitis.Methods:Rats were randomly divided into three groups:blank group,model group and acupoint group.The rat model of ovalbumin(OVA)AR was established,and the general condition of the rats was observed and scored.Acupuncture intervention was performed on the acupoint group on the second day after successful modeling,once per day for 20 min for 10 d.After intervention,the general behavior,behavioral score and histomorphological changes of nasal mucosa were observed.Eosinophils(EOS)were counted under microscope after nasal lavage smear staining,and the contents of total IgE,IFN-γ,IL-12,IL-4 and IL-5 in serum were detected by ELISA.Westernblot and IHC were used to detect the protein level and positive protein expression of specific transcription factors T-bet and GATA-3 in rat nasal mucosa.Results:After the establishment of the model,except for the blank group,the behavioral observation scores of rats in the model group and acupoint group were more than 5 points,indicating that the model was successful.After acupuncture intervention on acupoint"Yingxiang-Hegu",the behavioral score of rats in the acupoint group and western medicine group was significantly lower than that in the model group(P<0.05).Microscopic examination showed that the structure of nasal mucosa in the model group was obviously damaged,cilia were arranged discontinuously,uneven,local congestion and swelling,a large number of epithelial cells exfoliated and necrotic,goblet cell proliferation,obvious inflammatory cell infiltration.The pathological degree of nasal mucosa in the pair point group was significantly less than that in the model group.Compared with the model group,the levels of IFN-γ,IL-2 and IL-12 in serum were significantly increased,while IgE,IL-4,IL-5 and IL-6 were significantly decreased,GATA-3 protein and positive expression in nasal mucosa were significantly decreased and T-bet was significantly increased after acupuncture.Conclusion:Acupuncture at"Yingxiang-Hegu"can effectively improve the nasal sensitive symptoms and control nasal inflammation in AR rats.The mechanism may be that acupuncture at Yingxiang-Hegu can up-regulate the expression of T-bet,decrease the level of GATA-3,promote the production of Th1 cytokines and inhibit the synthesis and secretion of Th2 cells,thus restoring the immune balance of Th1 and Th2.

基于PointNet的钢板毛坯垛点云分割

钢板毛坯垛拆垛工序中,钢板毛坯分层厚度的估计是推钢机准确且安全执行推钢动作的关键;当前不少企业此道生产工序仍主要靠操作员人工观察的方式估算钢板厚度及与传送辊道的相对高度,其估算不准易导致碰撞,造成输送设备损坏生产中断。提出一种钢板毛坯垛智能分层方法,该方法结合现场工况环境采用激光雷达,对钢板毛坯垛进行三维点云成像,然后对采集的点云数据,用PointNet神经网络框架进行特征识别、分层分割与提取,最后对分割的不同层根据标定值,换算成真实厚度。根据现场实验结果表明,PointNet对钢板毛坯垛分割的识别率达到了87.4%,厚度估算误差低于1.2cm,根据钢厂钢板毛坯规格表(3种规格150、160、180cm)可以准确估计出钢板毛坯厚度的规格,识别速度15帧/s,满足现场工况要求。

基于轻量化PointNet网络的林果园喷雾作业靶标实时识别方法

为了进一步提高喷雾机器人靶标检测的精准性、实时性和应用部署的实用性,该研究提出一种基于轻量化PointNet网络的林果园喷雾作业靶标实时识别方法。首先通过区域提取降采样、地面分割和改进DBSCAN聚类等点云预处理方法提取原始点云中的靶标;然后通过移动最小二乘上采样将靶标点云转化为满足点云识别网络输入要求的点云数据;最终通过在PointNet网络中引入残差模块和改进循环剪枝算法轻量化PointNet网络,完成林果树靶标的实时识别。试验结果表明,在ModelNet40数据集上,轻量化PointNet网络可达89.7%的准确率;在实际苗圃环境的试验中,该研究方法对靶标的识别准确率可达92.49%,同时误识率与拒识率分别为13.4%和6.47%,相较PointNet网络识别准确率提升了4.38个百分点,误识率和拒识率分别降低了7.2和4.07个百分点;轻量化PointNet网络识别准确率仅比PointNet++网络低1.14个百分点,误识率和拒识率分别高了0.9和1.12个百分点。但是轻量化PointNet网络的模型参数量较PointNet网络和PointNet++网络的模型参数量显著减少,仅为PointNet网络的11.5%,PointNet++网络的27.02%;运算量相较PointNet网络、PointNet++网络分别减少13.3和76.79个百分点。该研究提出的轻量化PointNet网络具有较高的实时性、精确性和鲁棒性,能够满足林果园喷雾作业的靶标识别需求,可为林果园喷雾作业靶标实时识别提供参考。

Coexistence of Dirac and Weyl points in non-centrosymmetric semimetal NbIrTe_(4)

Using angle-resolved photoemission spectroscopy and density functional theory calculations methods,we investigate the electronic structures and topological properties of ternary tellurides NbIrTe_(4),a candidate for type-II Weyl semimetal.We demonstrate the presence of several Fermi arcs connecting their corresponding Weyl points on both termination surfaces of the topological material.Our analysis reveals the existence of Dirac points,in addition to Weyl points,giving both theoretical and experimental evidences of the coexistence of Dirac and Weyl points in a single material.These findings not only confirm NbIrTe_(4) as a unique topological semimetal but also open avenues for exploring novel electronic devices based on its coexisting Dirac and Weyl fermions.

Enhanced Mechanism for Link Failure Rerouting in Software-Defined Exchange Point Networks

Internet Exchange Point(IXP)is a system that increases network bandwidth performance.Internet exchange points facilitate interconnection among network providers,including Internet Service Providers(ISPs)andContent Delivery Providers(CDNs).To improve service management,Internet exchange point providers have adopted the Software Defined Network(SDN)paradigm.This implementation is known as a Software-Defined Exchange Point(SDX).It improves network providers’operations and management.However,performance issues still exist,particularly with multi-hop topologies.These issues include switch memory costs,packet processing latency,and link failure recovery delays.The paper proposes Enhanced Link Failure Rerouting(ELFR),an improved mechanism for rerouting link failures in software-defined exchange point networks.The proposed mechanism aims to minimize packet processing time for fast link failure recovery and enhance path calculation efficiency while reducing switch storage overhead by exploiting the Programming Protocol-independent Packet Processors(P4)features.The paper presents the proposed mechanisms’efficiency by utilizing advanced algorithms and demonstrating improved performance in packet processing speed,path calculation effectiveness,and switch storage management compared to current mechanisms.The proposed mechanism shows significant improvements,leading to a 37.5%decrease in Recovery Time(RT)and a 33.33%decrease in both Calculation Time(CT)and Computational Overhead(CO)when compared to current mechanisms.The study highlights the effectiveness and resource efficiency of the proposed mechanism in effectively resolving crucial issues inmulti-hop software-defined exchange point networks.

On a Heuristic Point of View about the Generalization of Curie Law to Cosmic Higgs Fields with the Casimir Effect

Condensed state physics demonstrates that the Curie temperature is the point at which spontaneous magnetization drops to zero, marking the critical transition where ferromagnetic or ferrimagnetic materials transform into paramagnetic substances. Below the Curie temperature, a material remains ferromagnetic;above it, the material becomes paramagnetic, with its magnetic field easily influenced by external magnetic fileds. For example, the Curie temperature of iron (Fe) is 1043 K, while that of neodymium magnets ranges from 583 to 673 K. From both physics and mathematics perspectives, examining the temperature properties of materials is essential, as it provides valuable insights into their electromagnetic and thermodynamic behaviors. This paper makes a bold assumption and, for the first time, carefully verifies the existence of a Casimir temperature at 0.00206 K under conditions of one-atomic spacing.

Qualitative Properties of Equilibrium Point in a Discrete Predator-Prey Model

In this paper, the dynamic properties of a discrete predator-prey model are discussed. The properties of non-hyperbolic fixed points and hyperbolic fixed points of the model are analyzed. First, by using the classic Shengjin formula, we find the existence conditions for fixed points of the model. Then, by using the qualitative theory of ordinary differential equations and matrix theory we indicate which points are hyperbolic and which are non-hyperbolic and the associated conditions.

Efficacy comparison of multipoint and single point scanning panretinal laser photocoagulation in non-proliferative diabetic retinopathy treatment

BACKGROUND Non-proliferative diabetic retinopathy(NPDR)poses a significant challenge in diabetes management due to its microvascular changes in the retina.Laser photocoagulation,a conventional therapy,aims to mitigate the risk of progressing to proliferative diabetic retinopathy(PDR).AIM To compare the efficacy and safety of multi-spot vs single-spot scanning panretinal laser photocoagulation in NPDR patients.METHODS Forty-nine NPDR patients(86 eyes)treated between September 2020 and July 2022 were included.They were randomly allocated into single-spot(n=23,40 eyes)and multi-spot(n=26,46 eyes)groups.Treatment outcomes,including bestcorrected visual acuity(BCVA),central macular thickness(CMT),and mean threshold sensitivity,were assessed at predetermined intervals over 12 months.Adverse reactions were also recorded.RESULTS Energy levels did not significantly differ between groups(P>0.05),but the multi-spot group exhibited lower energy density(P<0.05).BCVA and CMT improvements were noted in the multi-spot group at one-month posttreatment(P<0.05).Adverse reaction incidence was similar between groups(P>0.05).CONCLUSION While energy intensity and safety were comparable between modalities,multi-spot scanning demonstrated lower energy density and showed superior short-term improvements in BCVA and CMT for NPDR patients,with reduced laser-induced damage.

Building Facade Point Clouds Segmentation Based on Optimal Dual-Scale Feature Descriptors

To address the current issues of inaccurate segmentation and the limited applicability of segmentation methods for building facades in point clouds, we propose a facade segmentation algorithm based on optimal dual-scale feature descriptors. First, we select the optimal dual-scale descriptors from a range of feature descriptors. Next, we segment the facade according to the threshold value of the chosen optimal dual-scale descriptors. Finally, we use RANSAC (Random Sample Consensus) to fit the segmented surface and optimize the fitting result. Experimental results show that, compared to commonly used facade segmentation algorithms, the proposed method yields more accurate segmentation results, providing a robust data foundation for subsequent 3D model reconstruction of buildings.

Existence of Monotone Positive Solution for a Fourth-Order Three-Point BVP with Sign-Changing Green’s Function

This paper is concerned with the following fourth-order three-point boundary value problem , where , we discuss the existence of positive solutions to the above problem by applying to the fixed point theory in cones and iterative technique.

基于Point Transformer v2的点云枝叶分离方法研究

准确高效的点云枝叶分离对精确计算森林树木的垂直参数至关重要。然而,当前的研究方法计算成本高,且依赖先验知识导致泛化能力不足。针对以上问题,文章提出利用基于点特征的Transformer网络进行自动化的森林场景三维点云的枝叶分离研究。该方法使用Point Transformer v2网络,首先利用网格编码模块提取可学习的局部结构关系,保留点云的几何拓扑结构;其次使用分组注意力实现多通道联合学习,降低特征的冗余度,提高计算的效率;最后构建了基于点的Transformer网络实现高精度森林树木三维点云语义分割,降低了对于先验知识的需求。使用地基激光扫描仪获取的加拿大和芬兰7个不同树种样地的三维点云数据,进行枝叶分离实验和精度评价。实验结果表明,网络整体精度(OA)为94.42%,mIoU为78.89%,能够适应不同树种、不同点云密度的森林场景的枝叶分离。

基于改进PointNet++的船体分段合拢面构件智能识别算法研究

[目的]三维扫描仪获得的船体分段合拢面点云数据,具有精度高、数据量大的优势,能够很好地反映分段合拢面的建造状况。由于现有的PointNet++网络无法处理大容量点云数据,因此提出一种基于改进PointNet++的船体分段合拢面构件智能识别算法,实现针对大容量船体分段合拢面点云数据构件的智能识别。[方法]基于超体素生长理论对船体分段合拢面点云数据进行分割及简化,构建船体分段合拢面点云数据集,并使用该数据集训练基于深度学习理论改进的PointNet++网络。[结果]网络模型在船体分段合拢面点云数据训练集和测试集上的收敛结果趋于稳定,在测试集上识别准确率达到90.012%。[结论]该方法具有良好的识别能力,能够完成船体分段合拢面构件的智能识别。

Diagnosis of indirectly driven double shell targets with point-projection hard x-ray radiography

We present an application of short-pulse laser-generated hard x rays for the diagnosis of indirectly driven double shell targets. Coneinserted double shell targets were imploded through an indirect drive approach on the upgraded SG-II laser facility. Then, based on thepoint-projection hard x-ray radiography technique, time-resolved radiography of the double shell targets, including that of their near-peakcompression, were obtained. The backlighter source was created by the interactions of a high-intensity short pulsed laser with a metalmicrowire target. Images of the target near peak compression were obtained with an Au microwire. In addition, radiation hydrodynamicsimulations were performed, and the target evolution obtained agrees well with the experimental results. Using the radiographic images, arealdensities of the targets were evaluated.

Point Cloud Classification Using Content-Based Transformer via Clustering in Feature Space

Recently, there have been some attempts of Transformer in 3D point cloud classification. In order to reduce computations, most existing methods focus on local spatial attention,but ignore their content and fail to establish relationships between distant but relevant points. To overcome the limitation of local spatial attention, we propose a point content-based Transformer architecture, called PointConT for short. It exploits the locality of points in the feature space(content-based), which clusters the sampled points with similar features into the same class and computes the self-attention within each class, thus enabling an effective trade-off between capturing long-range dependencies and computational complexity. We further introduce an inception feature aggregator for point cloud classification, which uses parallel structures to aggregate high-frequency and low-frequency information in each branch separately. Extensive experiments show that our PointConT model achieves a remarkable performance on point cloud shape classification. Especially, our method exhibits 90.3% Top-1 accuracy on the hardest setting of ScanObjectN N. Source code of this paper is available at https://github.com/yahuiliu99/PointC onT.

Material point method simulation of hydro-mechanical behaviour in twophase porous geomaterials: A state-of-the-art review

The material point method(MPM)has been gaining increasing popularity as an appropriate approach to the solution of coupled hydro-mechanical problems involving large deformation.In this paper,we survey the current state-of-the-art in the MPM simulation of hydro-mechanical behaviour in two-phase porous geomaterials.The review covers the recent advances and developments in the MPM and their extensions to capture the coupled hydro-mechanical problems involving large deformations.The focus of this review is aiming at providing a clear picture of what has or has not been developed or implemented for simulating two-phase coupled large deformation problems,which will provide some direct reference for both practitioners and researchers.