Mesh representation matters:investigating the influence of different mesh features on perceptual and spatial fidelity of deep 3D morphable models

Background Deep 3D morphable models(deep 3DMMs)play an essential role in computer vision.They are used in facial synthesis,compression,reconstruction and animation,avatar creation,virtual try-on,facial recognition systems and medical imaging.These applications require high spatial and perceptual quality of synthesised meshes.Despite their significance,these models have not been compared with different mesh representations and evaluated jointly with point-wise distance and perceptual metrics.Methods We compare the influence of different mesh representation features to various deep 3DMMs on spatial and perceptual fidelity of the reconstructed meshes.This paper proves the hypothesis that building deep 3DMMs from meshes represented with global representations leads to lower spatial reconstruction error measured with L_(1) and L_(2) norm metrics and underperforms on perceptual metrics.In contrast,using differential mesh representations which describe differential surface properties yields lower perceptual FMPD and DAME and higher spatial fidelity error.The influence of mesh feature normalisation and standardisation is also compared and analysed from perceptual and spatial fidelity perspectives.Results The results presented in this paper provide guidance in selecting mesh representations to build deep 3DMMs accordingly to spatial and perceptual quality objectives and propose combinations of mesh representations and deep 3DMMs which improve either perceptual or spatial fidelity of existing methods.

Attention Guided Multi Scale Feature Fusion Network for Automatic Prostate Segmentation

The precise and automatic segmentation of prostate magnetic resonance imaging(MRI)images is vital for assisting doctors in diagnosing prostate diseases.In recent years,many advanced methods have been applied to prostate segmentation,but due to the variability caused by prostate diseases,automatic segmentation of the prostate presents significant challenges.In this paper,we propose an attention-guided multi-scale feature fusion network(AGMSF-Net)to segment prostate MRI images.We propose an attention mechanism for extracting multi-scale features,and introduce a 3D transformer module to enhance global feature representation by adding it during the transition phase from encoder to decoder.In the decoder stage,a feature fusion module is proposed to obtain global context information.We evaluate our model on MRI images of the prostate acquired from a local hospital.The relative volume difference(RVD)and dice similarity coefficient(DSC)between the results of automatic prostate segmentation and ground truth were 1.21%and 93.68%,respectively.To quantitatively evaluate prostate volume on MRI,which is of significant clinical significance,we propose a unique AGMSF-Net.The essential performance evaluation and validation experiments have demonstrated the effectiveness of our method in automatic prostate segmentation.

SGT-Net: A Transformer-Based Stratified Graph Convolutional Network for 3D Point Cloud Semantic Segmentation

In recent years,semantic segmentation on 3D point cloud data has attracted much attention.Unlike 2D images where pixels distribute regularly in the image domain,3D point clouds in non-Euclidean space are irregular and inherently sparse.Therefore,it is very difficult to extract long-range contexts and effectively aggregate local features for semantic segmentation in 3D point cloud space.Most current methods either focus on local feature aggregation or long-range context dependency,but fail to directly establish a global-local feature extractor to complete the point cloud semantic segmentation tasks.In this paper,we propose a Transformer-based stratified graph convolutional network(SGT-Net),which enlarges the effective receptive field and builds direct long-range dependency.Specifically,we first propose a novel dense-sparse sampling strategy that provides dense local vertices and sparse long-distance vertices for subsequent graph convolutional network(GCN).Secondly,we propose a multi-key self-attention mechanism based on the Transformer to further weight augmentation for crucial neighboring relationships and enlarge the effective receptive field.In addition,to further improve the efficiency of the network,we propose a similarity measurement module to determine whether the neighborhood near the center point is effective.We demonstrate the validity and superiority of our method on the S3DIS and ShapeNet datasets.Through ablation experiments and segmentation visualization,we verify that the SGT model can improve the performance of the point cloud semantic segmentation.

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.

Monocular 3D object detection with Pseudo-LiDAR confidence sampling and hierarchical geometric feature extraction in 6G network

The high bandwidth and low latency of 6G network technology enable the successful application of monocular 3D object detection on vehicle platforms.Monocular 3D-object-detection-based Pseudo-LiDAR is a low-cost,lowpower solution compared to LiDAR solutions in the field of autonomous driving.However,this technique has some problems,i.e.,(1)the poor quality of generated Pseudo-LiDAR point clouds resulting from the nonlinear error distribution of monocular depth estimation and(2)the weak representation capability of point cloud features due to the neglected global geometric structure features of point clouds existing in LiDAR-based 3D detection networks.Therefore,we proposed a Pseudo-LiDAR confidence sampling strategy and a hierarchical geometric feature extraction module for monocular 3D object detection.We first designed a point cloud confidence sampling strategy based on a 3D Gaussian distribution to assign small confidence to the points with great error in depth estimation and filter them out according to the confidence.Then,we present a hierarchical geometric feature extraction module by aggregating the local neighborhood features and a dual transformer to capture the global geometric features in the point cloud.Finally,our detection framework is based on Point-Voxel-RCNN(PV-RCNN)with high-quality Pseudo-LiDAR and enriched geometric features as input.From the experimental results,our method achieves satisfactory results in monocular 3D object detection.

Face recognition using SIFT features under 3D meshes

Expression, occlusion, and pose variations are three main challenges for 3D face recognition. A novel method is presented to address 3D face recognition using scale-invariant feature transform(SIFT) features on 3D meshes. After preprocessing, shape index extrema on the 3D facial surface are selected as keypoints in the difference scale space and the unstable keypoints are removed after two screening steps. Then, a local coordinate system for each keypoint is established by principal component analysis(PCA).Next, two local geometric features are extracted around each keypoint through the local coordinate system. Additionally, the features are augmented by the symmetrization according to the approximate left-right symmetry in human face. The proposed method is evaluated on the Bosphorus, BU-3DFE, and Gavab databases, respectively. Good results are achieved on these three datasets. As a result, the proposed method proves robust to facial expression variations, partial external occlusions and large pose changes.

A Novel Airborne 3D Laser Point Cloud Hole Repair Algorithm Considering Topographic Features

Hole repair processing is an important part of point cloud data processing in airborne 3-dimensional(3D)laser scanning technology.Due to the fragmentation and irregularity of the surface morphology,when applying the 3D laser scanning technology to mountain mapping,the conventional mathematical cloud-based point cloud hole repair method is not ideal in practical applications.In order to solve this problem,we propose to repair the valley and ridge line first,and then repair the point cloud hole.The main technical steps of the method include the following points:First,the valley and ridge feature lines are extracted by the GIS slope analysis method;Then,the valley and ridge line missing from the hole are repaired by the mathematical interpolation method,and the repaired results are edited and inserted to the original point cloud;Finally,the traditional repair method is used to repair the point cloud hole whose valley line and ridge line have been repaired.Three experiments were designed and implemented in the east bank of the Xiaobaini River to test the performance of the proposed method.The results showed that compared with the direct point cloud hole repair method in Geomagic Studio software,the average repair accuracy of the proposed method,in the 16 m buffer zone of valley line and ridge line,is increased from 56.31 cm to 31.49 cm.The repair performance is significantly improved.

基于改进DETR的机器人铆接缺陷检测方法研究

铆接作为铁道车辆结构件的主要连接方式,合格的铆接质量是车辆安全稳定运行的重要保证。针对现有铆接缺陷检测方法存在检测精度低、检测点位少、检测智能化水平不高等问题,提出一种基于改进DETR的机器人铆接缺陷检测方法。首先,搭建铆接缺陷检测系统,依次采集工件尺寸大、铆钉尺寸小工况下的铆接缺陷图像。其次,为了增强DETR模型在小目标中的图像特征提取能力和检测性能,以EfficientNet作为DETR中的主干特征提取网络,并将3-D权重注意力机制SimAM引入EfficientNet网络,从而有效保留图像特征层的镦头形态信息和铆点区域的空间信息。然后,在颈部网络中引入加权双向特征金字塔模块,以EfficientNet网络的输出作为特征融合模块的输入对各尺度特征信息进行聚合,增大不同铆接缺陷的类间差异。最后,利用Smooth L1和DIoU的线性组合改进原模型预测网络的回归损失函数,提高模型的检测精度和收敛速度。结果表明,改进模型表现出较高的检测性能,对于铆接缺陷的平均检测精度mAP为97.12%,检测速度FPS为25.4帧/s,与Faster RCNN、YOLOX等其他主流检测模型相比,在检测精度和检测速度方面均具有较大优势。研究结果能够满足实际工况中大型铆接件的小尺寸铆钉铆接缺陷实时在线检测的需求,为视觉检测技术在铆接工艺中的应用提供一定的参考价值。

急性脑梗死患者血清CTRP-3、D-二聚体、sTREM2水平及相关临床特征与溶栓后出血性转化的关系

目的 探讨急性脑梗死患者血清补体C1q/肿瘤坏死因子相关蛋白3(CTRP-3)、D-二聚体、可溶性髓样细胞触发受体2(sTREM2)水平及相关临床特征与溶栓后出血性转化(HT)的关系。方法 回顾性分析2018年9月—2022年9月在青海省人民医院接受溶栓治疗的120例急性脑梗死患者的临床资料,根据患者溶栓后是否发生HT分为HT组(30例)、非HT组(90例)。比较两组患者的临床资料及血清CTRP-3、D-二聚体、sTREM2水平。采用多因素逐步Logistic回归分析急性脑梗死患者溶栓后发生HT的危险因素;绘制受试者工作特征(ROC)曲线,分析急性脑梗死患者溶栓后HT预测模型预测HT发生的价值。结果 HT组心房颤动(以下简称房颤)、大面积脑梗死、入院NIHSS评分≥15分占比高于非HT组(P <0.05),血清CTRP-3水平低于非HT组(P <0.05),D-二聚体、sTREM2水平高于非HT组(P <0.05)。血清CTRP-3、D-二聚体、sTREM2水平预测急性脑梗死患者溶栓后发生HT的敏感性分别为66.7%(95%CI:0.598,0.756)、70.0%(95%CI:0.607,0.812)、80.0%(95%CI:0.714,0.889),特异性分别为73.3%(95%CI:0.636,0.821)、86.7%(95%CI:0.778,0.923)、86.7%(95%CI:0.747,0.942)。多因素Logistic逐步回归分析结果显示,房颤[OR=1.237(95%CI:1.103,1.387)]、大面积脑梗死[OR=2.338(95%CI:1.292,4.231)]、入院NIHSS评分≥15分[OR=2.087(95%CI:1.231,3.538)]、CTRP-3≤269.265μg/L [OR=3.006(95%CI:1.508,5.992)]、D-二聚体≥2.625 mg/L [OR=2.649(95%CI:1.374,5.107)]、sTREM2≥314.675 ng/L [OR=2.328(95%CI:1.411,3.841)]是急性脑梗死患者溶栓后发生HT的危险因素(P <0.05)。根据多因素Logistic逐步回归分析结果建立急性脑梗死患者溶栓后HT预测模型,Logit(P)=-33.887+0.213×房颤+0.849×大面积脑梗死+0.736×入院NIHSS评分+1.101×CTRP-3+0.974×D-二聚体+0.845×sTREM2;ROC曲线分析结果表明,预测模型预测HT发生的敏感性为93.3%(95%CI:0.841,0.991),特异性为87.8%(95%CI:0.808,0.976)。结论 血清CTRP-3、D-二聚体、sTREM2水平与急性脑梗死患者溶栓后HT有关,预测价值较高,且急性脑梗死患者溶栓后HT预测模型预测HT优于各项指标单独预测。

General and robust voxel feature learning with Transformer for 3D object detection

The self-attention networks and Transformer have dominated machine translation and natural language processing fields,and shown great potential in image vision tasks such as image classification and object detection.Inspired by the great progress of Transformer,we propose a novel general and robust voxel feature encoder for 3D object detection based on the traditional Transformer.We first investigate the permutation invariance of sequence data of the self-attention and apply it to point cloud processing.Then we construct a voxel feature layer based on the self-attention to adaptively learn local and robust context of a voxel according to the spatial relationship and context information exchanging between all points within the voxel.Lastly,we construct a general voxel feature learning framework with the voxel feature layer as the core for 3D object detection.The voxel feature with Transformer(VFT)can be plugged into any other voxel-based 3D object detection framework easily,and serves as the backbone for voxel feature extractor.Experiments results on the KITTI dataset demonstrate that our method achieves the state-of-the-art performance on 3D object detection.

6DOF pose estimation of a 3D rigid object based on edge-enhanced point pair features

The point pair feature(PPF)is widely used for 6D pose estimation.In this paper,we propose an efficient 6D pose estimation method based on the PPF framework.We introduce a well-targeted down-sampling strategy that focuses on edge areas for efficient feature extraction for complex geometry.A pose hypothesis validation approach is proposed to resolve ambiguity due to symmetry by calculating the edge matching degree.We perform evaluations on two challenging datasets and one real-world collected dataset,demonstrating the superiority of our method for pose estimation for geometrically complex,occluded,symmetrical objects.We further validate our method by applying it to simulated punctures.

Research on the relationship between geophysical structural features and earthquakes in Mid-Yunnan and the surrounding area

In this study, we analyzed the gravity and, magnetic characteristics, and the occurrence of a fault zone and discussed the relationships between the two locations. The results reveal that the subsurface structures strikes are different compared with those in the research region. In other words, the geophysical advantageous directions from the gravity and magnetic anomalies are not the same as those caused by the surface structures. The local horizontal gradient results from the gravity and magnetic anomalies show that the majority of earthquakes occur along an intense fault zone, which is a zone of abrupt gravity and negative magnetic change, where the shapes match very well. From the distribution of earthquakes in this area, we find that it has experienced more than 11 earthquake events with magnitude larger than Ms7.0. In addition, water development sites such as Jinshajiang, Lancangjiang, and the Red River and Pearl River watersheds have been hit ten times by earthquakes of this magnitude. It is observed that strong earthquakes occur frequently in the Holocene active fault zone.

3D simulation of image-defined complex internal features using the numerical manifold method

The numerical simulation of internal features,such as inclusions and voids,is important to analyze their impact on the performance of composite materials.However,the complex geometries of internal features and the induced continuous-discontinuous(C-D)deformation fields are challenges to their numerical simulation.In this study,a 3D approach using a simple mesh to simulate irregular internal geometries is developed for the first time.With the help of a developed voxel crack model,image models that are efficient when recording complex geometries are directly imported into the simulation.Surface reconstructions,which are usually labor-intensive,are excluded from this approach.Moreover,using image models as the geometric input,image processing techniques are applied to detect material interfaces and develop contact pairs.Then,the C-D deformations of the complex internal features are directly calculated based on the numerical manifold method.The accuracy and convergence of the developed3D approach are examined based on multiple benchmarks.Successful 3D C-D simulation of sandstones with naturally formed complex microfeatures demonstrates the capability of the developed approach.

Movement Function Assessment Based on Human Pose Estimation from Multi-View

Human pose estimation is a basic and critical task in the field of computer vision that involves determining the position(or spatial coordinates)of the joints of the human body in a given image or video.It is widely used in motion analysis,medical evaluation,and behavior monitoring.In this paper,the authors propose a method for multi-view human pose estimation.Two image sensors were placed orthogonally with respect to each other to capture the pose of the subject as they moved,and this yielded accurate and comprehensive results of three-dimensional(3D)motion reconstruction that helped capture their multi-directional poses.Following this,we propose a method based on 3D pose estimation to assess the similarity of the features of motion of patients with motor dysfunction by comparing differences between their range of motion and that of normal subjects.We converted these differences into Fugl–Meyer assessment(FMA)scores in order to quantify them.Finally,we implemented the proposed method in the Unity framework,and built a Virtual Reality platform that provides users with human–computer interaction to make the task more enjoyable for them and ensure their active participation in the assessment process.The goal is to provide a suitable means of assessing movement disorders without requiring the immediate supervision of a physician.

Advanced 3D ordered electrodes for PEMFC applications: From structural features and fabrication methods to the controllable design of catalyst layers

The catalyst layers(CLs) electrode is the key component of the membrane electrode assembly(MEA) in proton exchange membrane fuel cells(PEMFCs). Conventional electrodes for PEMFCs are composed of carbon-supported, ionomer, and Pt nanoparticles, all immersed together and sprayed with a micron-level thickness of CLs. They have a performance trade-off where increasing the Pt loading leads to higher performance of abundant triple-phase boundary areas but increases the electrode cost. Major challenges must be overcome before realizing its wide commercialization. Literature research revealed that it is impossible to achieve performance and durability targets with only high-performance catalysts, so the controllable design of CLs architecture in MEAs for PEMFCs must now be the top priority to meet industry goals. From this perspective, a 3D ordered electrode circumvents this issue with a support-free architecture and ultrathin thickness while reducing noble metal Pt loadings. Herein, we discuss the motivation in-depth and summarize the necessary CLs structural features for designing ultralow Pt loading electrodes. Critical issues that remain in progress for 3D ordered CLs must be studied and characterized. Furthermore, approaches for 3D ordered CLs architecture electrode development, involving material design, structure optimization, preparation technology, and characterization techniques, are summarized and are expected to be next-generation CLs for PEMFCs. Finally, the review concludes with perspectives on possible research directions of CL architecture to address the significant challenges in the future.

(1,3)-β-D-葡聚糖在不同证据级别的侵袭性肺曲霉病中的表现

目的分析(1,3)-β-D-葡聚糖实验(G实验)在不同证据级别的侵袭性肺曲霉病中的表现。方法选取2014年1月-2017年12月在广东省人民医院诊断为侵袭性肺曲霉病的患者病例信息,收集患者临床信息:人口学资料、肺穿刺组织病理结果、呼吸道标本(痰、纤支镜冲洗液、肺泡灌洗液)培养结果、G实验结果。根据病理和培养的结果分为3组:病理+培养双阳性组、单培养阳性组、单病理阳性组。结果最后纳入分析的病例76例,男性46人,女性30人,平均年龄59±21岁。分组结果为:单病理阳性组39例(41.5%),病理+培养双阳性组10例(10.6%),单培养阳性组45例(47.9%)。3组的G实验阳性率有统计学差异(χ^2=9.34,P=0.036),进行两两比较发现,a调整为0.0125后,双阳性组的G实验阳性率显著高于单病理阳性组(χ^2=9.06,P1=0.004),G实验的中位值也是双阳性组最高(249.30pg/ml)。呼吸道标本培养最常见的是烟曲霉36例(占61.02%),其次是黄曲霉12例(占20.34%),但不同曲霉感染其G实验结果无统计学差异(H=4.021,P=0.403)。结论病理阳性的侵袭性肺曲霉患者其G实验的结果不一定高,最好结合培养结果进行诊断。

基于特征点的3D人脸姿态跟踪

针对视频序列中的人脸跟踪问题,提出一种单摄像头的人脸3D姿态跟踪方法。利用SIFT特征点匹配算法来得到可靠的帧间特征匹配。将前帧与所选的关键帧特征匹配信息及融入到对当前的姿态估计中,利用SIFT特征点匹配算法来得到可靠的帧间特征匹配。最后通过利用RANSAC随机选取特征点对,并用POSIT和最小化误差组合的3D投影方法以迭代的方式得到精确的当前帧人脸姿态估计。通过多组实验数据对比,表明了该算法在严重遮挡、头部摆动幅度较大、匹配点较少的复杂情况干扰下仍具有鲁棒性,并且解决了3D人脸跟踪的漂移问题,实现对目标人脸的稳定跟踪,对比以往2D跟踪算法在复杂环境下具有明显的改善。

A Novel Human Action Recognition Algorithm Based on Decision Level Multi-Feature Fusion

In order to take advantage of the logical structure of video sequences and improve the recognition accuracy of the human action, a novel hybrid human action detection method based on three descriptors and decision level fusion is proposed. Firstly, the minimal 3D space region of human action region is detected by combining frame difference method and Vi BE algorithm, and the three-dimensional histogram of oriented gradient(HOG3D) is extracted. At the same time, the characteristics of global descriptors based on frequency domain filtering(FDF) and the local descriptors based on spatial-temporal interest points(STIP) are extracted. Principal component analysis(PCA) is implemented to reduce the dimension of the gradient histogram and the global descriptor, and bag of words(BoW) model is applied to describe the local descriptors based on STIP. Finally, a linear support vector machine(SVM) is used to create a new decision level fusion classifier. Some experiments are done to verify the performance of the multi-features, and the results show that they have good representation ability and generalization ability. Otherwise, the proposed scheme obtains very competitive results on the well-known datasets in terms of mean average precision.

Fast and Stable Surface Feature Simulation for Particle-Based Fluids

In order to efficiently and realistically capture microscopic features of fluid surface,a fast and stable surface feature simulation approach for particle-based fluids is presented in this paper.This method employs a steady tension and adhesion model to construct surface features with the consideration of the adsorption effect of fluid to solid.Molecular cohesion and surface area minimization are appended for surface tension,and adhesion is added to better show the microscopic characteristics of fluid surface.Besides,the model is integrated to an implicit incompressible smoothed particle hydrodynamics(SPH)method to improve the efficiency and stability of simulation.The experimental results demonstrate that the method can better simulates surface features in a variety of scenarios stably and efficiently.

A fast, accurate and dense feature matching algorithm for aerial images

Three-dimensional(3D)reconstruction based on aerial images has broad prospects,and feature matching is an important step of it.However,for high-resolution aerial images,there are usually problems such as long time,mismatching and sparse feature pairs using traditional algorithms.Therefore,an algorithm is proposed to realize fast,accurate and dense feature matching.The algorithm consists of four steps.Firstly,we achieve a balance between the feature matching time and the number of matching pairs by appropriately reducing the image resolution.Secondly,to realize further screening of the mismatches,a feature screening algorithm based on similarity judgment or local optimization is proposed.Thirdly,to make the algorithm more widely applicable,we combine the results of different algorithms to get dense results.Finally,all matching feature pairs in the low-resolution images are restored to the original images.Comparisons between the original algorithms and our algorithm show that the proposed algorithm can effectively reduce the matching time,screen out the mismatches,and improve the number of matches.