Road accident detection plays an important role in abnormal scene reconstruction for Intelligent Transportation Systems and abnormal events warning for autonomous driving.This paper presents a novel 3D object detector...Road accident detection plays an important role in abnormal scene reconstruction for Intelligent Transportation Systems and abnormal events warning for autonomous driving.This paper presents a novel 3D object detector and adaptive space partitioning algorithm to infer traffic accidents quantitatively.Using 2D region proposals in an RGB image,this method generates deformable frustums based on point cloud for each 2D region proposal and then frustum-wisely extracts features based on the farthest point sampling network(FPS-Net)and feature extraction network(FE-Net).Subsequently,the encoder-decoder network(ED-Net)implements 3D-oriented bounding box(OBB)regression.Meanwhile,the adaptive least square regression(ALSR)method is proposed to split 3D OBB.Finally,the reduced OBB intersection test is carried out to detect traffic accidents via separating surface theorem(SST).In the experiments of KITTI benchmark,our proposed 3D object detector outperforms other state-of-theartmethods.Meanwhile,collision detection algorithm achieves the satisfactory performance of 91.8%accuracy on our SHTA dataset.展开更多
In this paper an evaluation of the influence of luminance L* at the L*a*b* color space during color segmentation is presented. A comparative study is made between the behavior of segmentation in color images using onl...In this paper an evaluation of the influence of luminance L* at the L*a*b* color space during color segmentation is presented. A comparative study is made between the behavior of segmentation in color images using only the Euclidean metric of a* and b* and an adaptive color similarity function defined as a product of Gaussian functions in a modified HSI color space. For the evaluation synthetic images were particularly designed to accurately assess the performance of the color segmentation. The testing system can be used either to explore the behavior of a similarity function (or metric) in different color spaces or to explore different metrics (or similarity functions) in the same color space. From the results is obtained that the color parameters a* and b* are not independent of the luminance parameter L* as one might initially assume.展开更多
北斗三号全球卫星导航系统(BDS-3)于2020年7月31日正式开通,其空间段服务能力是决定系统整体性能表现的重要因素.本文对广播轨道、广播钟差、空间信号测距误差、广播电离层精度的评估计算方法进行了分析,分别以GFZ(German Research Cent...北斗三号全球卫星导航系统(BDS-3)于2020年7月31日正式开通,其空间段服务能力是决定系统整体性能表现的重要因素.本文对广播轨道、广播钟差、空间信号测距误差、广播电离层精度的评估计算方法进行了分析,分别以GFZ(German Research Centre for Geosciences)和iGMAS(International GNSS Monitoring and Assessment System)的最终产品为参考基准,对系统从2020年正式开通至2025年的变化情况进行了评估分析.研究表明, BDS-3广播轨道精度呈现明显的卫星类型相关性, MEO卫星高于IGSO卫星,与GFZ产品和iGMAS产品相比,径向、切向、法向95%的RMS值均得到不同程度地改善;广播钟差误差与SISRE(Signal-InSpace Range Error) 95%的RMS值则提升了分米级的精度;电离层模型误差方面,在评估周期内, Klobuchar模型的VTEC值分布范围相对较广,低VTEC区间BDGIM模型的频次分布更为集中;与CODE和iGMAS电离层产品相比,在太阳活动极小期的2020年BDGIM(Beidou Global Lonospheric Delay Correction Model)模型VTEC平均RMS值优于Klobuchar模型,而在太阳活动极大期的2025年两模型的VTEC平均RMS值均呈上升的趋势,且BDGIM模型的稳定性更强.展开更多
目的随着电影内容的复杂化与多样化,电影场景分割成为理解影片结构和支持多媒体应用的重要任务。为提升镜头特征提取和特征关联的有效性,增强镜头序列的上下文感知能力,提出一种混合架构电影场景分割方法(hybrid architecture scene seg...目的随着电影内容的复杂化与多样化,电影场景分割成为理解影片结构和支持多媒体应用的重要任务。为提升镜头特征提取和特征关联的有效性,增强镜头序列的上下文感知能力,提出一种混合架构电影场景分割方法(hybrid architecture scene segmentation network,HASSNet)。方法首先,采用预训练结合微调策略,在大量无场景标签的电影数据上进行无监督预训练,使模型学习有效的镜头特征表示和关联特性,然后在有场景标签的数据上进行微调训练,进一步提升模型性能;其次,模型架构上混合了状态空间模型和自注意力机制模型,分别设计Shot Mamba镜头特征提取模块和Scene Transformer特征关联模块,Shot Mamba通过对镜头图像分块建模提取有效特征表示,Scene Transformer则通过注意力机制对不同镜头特征进行关联建模;最后,采用3种无监督损失函数进行预训练,提升模型在镜头特征提取和关联上的性能,并使用Focal Loss损失函数进行微调,以改善由于类别不平衡导致的精度不足问题。结果实验结果表明,HASSNet在3个数据集上显著提升了场景分割的精度,在典型电影场景分割数据集MovieNet中,与先进的场景分割方法相比,AP(average precision)、mIoU(mean intersection over union)、AUC-ROC(area under the receiver operating characteristic curve)和F1分别提升1.66%、10.54%、0.21%和16.83%,验证了本文提出的HASSNet方法可以有效提升场景边界定位的准确性。结论本文提出的HASSNet方法有效结合了预训练与微调策略,借助混合状态空间模型和自注意力机制模型的特点,增强了镜头的上下文感知能力,使电影场景分割的结果更加准确。展开更多
The geometry of Teichmuller metric in an asymptotic Teichmuller space is studled in this article. First, a binary infinitesimal form of Teichmuller metric on AT（X） is proved. Then, the notion of angles between two g...The geometry of Teichmuller metric in an asymptotic Teichmuller space is studled in this article. First, a binary infinitesimal form of Teichmuller metric on AT（X） is proved. Then, the notion of angles between two geodesic curves in the asymptotic Teichmuller space AT（X） is introduced. The existence of such angles is proved and the explicit formula is obtained. As an application, a sufficient condition for non-uniqueness geodesics in AT（X） is obtained.展开更多
Minimally Invasive Spine surgery (MISS) was developed to treat disorders of the spine with less disruption to the muscles. Surgeons use CT images to monitor the volume of muscles after operation in order to evaluate t...Minimally Invasive Spine surgery (MISS) was developed to treat disorders of the spine with less disruption to the muscles. Surgeons use CT images to monitor the volume of muscles after operation in order to evaluate the progress of patient recovery. The first step in the task is to segment the muscle regions from other tissues/organs in CT images. However, manual segmentation of muscle regions is not only inaccurate, but also time consuming. In this work, Gray Space Map (GSM) is used in fuzzy c-means clustering algorithm to segment muscle regions in CT images. GSM com- bines both spatial and intensity information of pixels. Experiments show that the proposed GSM- based fuzzy c-means clustering muscle CT image segmentation yields very good results.展开更多
基金National Natural Science Foundation of China(No.51805312)in part by Shanghai Sailing Program(No.18YF1409400)+4 种基金in part by Training and Funding Program of Shanghai College young teachers(No.ZZGCD15102)in part by Scientific Research Project of Shanghai University of Engineering Science(No.2016-19)in part by Science and Technology Commission of Shanghai Municipality(No.19030501100)in part by the Shanghai University of Engineering Science Innovation Fund for Graduate Students(No.18KY0613)in part by National Key R&D Program of China(No.2016YFC0802900).
文摘Road accident detection plays an important role in abnormal scene reconstruction for Intelligent Transportation Systems and abnormal events warning for autonomous driving.This paper presents a novel 3D object detector and adaptive space partitioning algorithm to infer traffic accidents quantitatively.Using 2D region proposals in an RGB image,this method generates deformable frustums based on point cloud for each 2D region proposal and then frustum-wisely extracts features based on the farthest point sampling network(FPS-Net)and feature extraction network(FE-Net).Subsequently,the encoder-decoder network(ED-Net)implements 3D-oriented bounding box(OBB)regression.Meanwhile,the adaptive least square regression(ALSR)method is proposed to split 3D OBB.Finally,the reduced OBB intersection test is carried out to detect traffic accidents via separating surface theorem(SST).In the experiments of KITTI benchmark,our proposed 3D object detector outperforms other state-of-theartmethods.Meanwhile,collision detection algorithm achieves the satisfactory performance of 91.8%accuracy on our SHTA dataset.
文摘In this paper an evaluation of the influence of luminance L* at the L*a*b* color space during color segmentation is presented. A comparative study is made between the behavior of segmentation in color images using only the Euclidean metric of a* and b* and an adaptive color similarity function defined as a product of Gaussian functions in a modified HSI color space. For the evaluation synthetic images were particularly designed to accurately assess the performance of the color segmentation. The testing system can be used either to explore the behavior of a similarity function (or metric) in different color spaces or to explore different metrics (or similarity functions) in the same color space. From the results is obtained that the color parameters a* and b* are not independent of the luminance parameter L* as one might initially assume.
文摘北斗三号全球卫星导航系统(BDS-3)于2020年7月31日正式开通,其空间段服务能力是决定系统整体性能表现的重要因素.本文对广播轨道、广播钟差、空间信号测距误差、广播电离层精度的评估计算方法进行了分析,分别以GFZ(German Research Centre for Geosciences)和iGMAS(International GNSS Monitoring and Assessment System)的最终产品为参考基准,对系统从2020年正式开通至2025年的变化情况进行了评估分析.研究表明, BDS-3广播轨道精度呈现明显的卫星类型相关性, MEO卫星高于IGSO卫星,与GFZ产品和iGMAS产品相比,径向、切向、法向95%的RMS值均得到不同程度地改善;广播钟差误差与SISRE(Signal-InSpace Range Error) 95%的RMS值则提升了分米级的精度;电离层模型误差方面,在评估周期内, Klobuchar模型的VTEC值分布范围相对较广,低VTEC区间BDGIM模型的频次分布更为集中;与CODE和iGMAS电离层产品相比,在太阳活动极小期的2020年BDGIM(Beidou Global Lonospheric Delay Correction Model)模型VTEC平均RMS值优于Klobuchar模型,而在太阳活动极大期的2025年两模型的VTEC平均RMS值均呈上升的趋势,且BDGIM模型的稳定性更强.
文摘目的随着电影内容的复杂化与多样化,电影场景分割成为理解影片结构和支持多媒体应用的重要任务。为提升镜头特征提取和特征关联的有效性,增强镜头序列的上下文感知能力,提出一种混合架构电影场景分割方法(hybrid architecture scene segmentation network,HASSNet)。方法首先,采用预训练结合微调策略,在大量无场景标签的电影数据上进行无监督预训练,使模型学习有效的镜头特征表示和关联特性,然后在有场景标签的数据上进行微调训练,进一步提升模型性能;其次,模型架构上混合了状态空间模型和自注意力机制模型,分别设计Shot Mamba镜头特征提取模块和Scene Transformer特征关联模块,Shot Mamba通过对镜头图像分块建模提取有效特征表示,Scene Transformer则通过注意力机制对不同镜头特征进行关联建模;最后,采用3种无监督损失函数进行预训练,提升模型在镜头特征提取和关联上的性能,并使用Focal Loss损失函数进行微调,以改善由于类别不平衡导致的精度不足问题。结果实验结果表明,HASSNet在3个数据集上显著提升了场景分割的精度,在典型电影场景分割数据集MovieNet中,与先进的场景分割方法相比,AP(average precision)、mIoU(mean intersection over union)、AUC-ROC(area under the receiver operating characteristic curve)和F1分别提升1.66%、10.54%、0.21%和16.83%,验证了本文提出的HASSNet方法可以有效提升场景边界定位的准确性。结论本文提出的HASSNet方法有效结合了预训练与微调策略,借助混合状态空间模型和自注意力机制模型的特点,增强了镜头的上下文感知能力,使电影场景分割的结果更加准确。
基金supported by National Natural Science Foundation of China(11371045,11301248)
文摘The geometry of Teichmuller metric in an asymptotic Teichmuller space is studled in this article. First, a binary infinitesimal form of Teichmuller metric on AT（X） is proved. Then, the notion of angles between two geodesic curves in the asymptotic Teichmuller space AT（X） is introduced. The existence of such angles is proved and the explicit formula is obtained. As an application, a sufficient condition for non-uniqueness geodesics in AT（X） is obtained.
文摘Minimally Invasive Spine surgery (MISS) was developed to treat disorders of the spine with less disruption to the muscles. Surgeons use CT images to monitor the volume of muscles after operation in order to evaluate the progress of patient recovery. The first step in the task is to segment the muscle regions from other tissues/organs in CT images. However, manual segmentation of muscle regions is not only inaccurate, but also time consuming. In this work, Gray Space Map (GSM) is used in fuzzy c-means clustering algorithm to segment muscle regions in CT images. GSM com- bines both spatial and intensity information of pixels. Experiments show that the proposed GSM- based fuzzy c-means clustering muscle CT image segmentation yields very good results.
