2.793

2018影响因子

(CJCR)

  • 中文核心
  • EI
  • 中国科技核心
  • Scopus
  • CSCD
  • 英国科学文摘

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

平行点云: 虚实互动的点云生成与三维模型进化方法

田永林 沈宇 李强 王飞跃

田永林, 沈宇, 李强, 王飞跃. 平行点云: 虚实互动的点云生成与三维模型进化方法. 自动化学报, 2020, 46(12): 2572−2582 doi: 10.16383/j.aas.c200800
引用本文: 田永林, 沈宇, 李强, 王飞跃. 平行点云: 虚实互动的点云生成与三维模型进化方法. 自动化学报, 2020, 46(12): 2572−2582 doi: 10.16383/j.aas.c200800
Tian Yong-Lin, Shen Yu, Li Qiang, Wang Fei-Yue. Parallel point clouds: point clouds generation and 3D model evolution via virtual-real interaction. Acta Automatica Sinica, 2020, 46(12): 2572−2582 doi: 10.16383/j.aas.c200800
Citation: Tian Yong-Lin, Shen Yu, Li Qiang, Wang Fei-Yue. Parallel point clouds: point clouds generation and 3D model evolution via virtual-real interaction. Acta Automatica Sinica, 2020, 46(12): 2572−2582 doi: 10.16383/j.aas.c200800

平行点云: 虚实互动的点云生成与三维模型进化方法

doi: 10.16383/j.aas.c200800
基金项目: 国家自然科学基金重点项目(61533019), 英特尔智能网联汽车大学合作研究中心项目(ICRI – IACV), 国家自然科学基金项目联合基金(U1811463), 广州市智能网联汽车重大科技专项(202007050002)资助
详细信息
    作者简介:

    田永林:中国科学技术大学与中国科学院自动化研究所联合培养博士研究生. 2017年获得中国科学技术大学自动化系学士学位. 主要研究方向为计算机视觉, 智能交通. E-mail: tyldyx@mail.ustc.edu.cn

    沈宇:中国科学院大学人工智能学院博士研究生. 主要研究方向为视频预测, 平行强化学习, 机器人自主导航. E-mail: shenyu2015@ia.ac.cn

    李强:中国科学院大学人工智能学院硕士研究生. 主要研究方向为图像识别, 平行无人车系统. E-mail: liqiang2018@ia.ac.cn

    王飞跃:中国科学院自动化研究所复杂系统管理与控制国家重点实验室研究员. 主要研究方向为智能系统和复杂系统的建模、分析与控制. 本文通信作者. E-mail: feiyue.wang@ia.ac.cn

Parallel Point Clouds: Point Clouds Generation and 3D Model Evolution via Virtual-real Interaction

Funds: Supported by the State Key Program of National Natural Science Foundation of China (61533019), the Intel Collaborative Research Institute for Intelligent and Automated Connected Vehicles (ICRI–IACV), the Joint Funds of the National Natural Science Foundation of China (U1811463), and the Key Research and Development Program of Guangzhou (202007050002)
  • 摘要: 三维信息的提取在自动驾驶等智能交通场景中正发挥着越来越重要的作用, 为了解决以激光雷达为主的深度传感器在数据采集方面面临的成本高、样本覆盖不全面等问题, 本文提出了平行点云的框架. 利用人工定义场景获取虚拟点云数据, 通过计算实验训练三维模型, 借助平行执行对模型性能进行测试, 并将结果反馈至数据生成和模型训练过程. 通过不断地迭代, 使三维模型得到充分评估并不断进化. 在平行点云的框架下, 我们以三维目标检测为例, 通过闭环迭代, 构建了虚实结合的点云数据集, 在无需人工标注的情况下, 可达到标注数据训练模型精度的72%.
  • 图  1  平行点云框架

    Fig.  1  The framework of parallel point clouds

    图  2  激光雷达扫描示意图

    Fig.  2  The illustration of LiDAR scanning

    图  3  CARLA仿真环境及虚拟雷达点云

    Fig.  3  CARLA simulation environment and virtual LiDAR point clouds

    图  4  KITTI伪雷达点云生成方法

    Fig.  4  Pesudo LiDAR point clouds generation with KITTI dataset

    图  5  虚实结合的点云生成方法

    Fig.  5  LiDAR point clouds generation via virtual-real interaction

    图  6  KITTI (左)与ShapeKITTI (右)点云数据可视化对比

    Fig.  6  The visualization of point clouds from KITTI (left) and ShapeKITTI (right)

    图  7  点云数据生成对三维检测模型性能的影响

    Fig.  7  The influence of point clouds generation on the 3D detector performance

    图  8  不同距离下的虚拟点云生成效果

    Fig.  8  The visualization of virtual point clouds under difierent distance

    图  9  不同方位角的虚拟点云生成效果

    Fig.  9  The visualization of virtual point clouds under difierent azimuths

    图  10  KITTI数据集中的长尾分布

    Fig.  10  The long-tailed distribution in KITTI dataset

    表  1  激光雷达属性描述

    Table  1  The description of LiDAR attributes

    属性描述默认值
    Channels激光器数量32
    Range测量/射线广播的
    最大距离(米)
    10.0
    Points per second每秒所有激光
    器产生的点
    56000
    Rotation frequency激光雷达旋转频率10.0
    Upper field of view最高激光的角度(度)10.0
    Lower field of view最低激光的角度(度)−30
    Atmosphere attenuation rate测量每米LiDAR强
    度损失的系数
    0.004
    Dropoff general rate随机掉落的总点数比例0.45
    Dropoff intensity limit对于基于强度的下降,阈值
    强度值不超过任何点
    0.8
    下载: 导出CSV

    表  2  基于ShapeKITTI的三维目标检测模型(PointPillars)平均精度 (%)

    Table  2  The average precision of 3D object detector (PointPillars) based on ShapeKITTI dataset (%)

    评测项目简单模式中等模式困难模式
    3D AP (Real)85.2175.8669.21
    3D AP (Ours)71.4454.3952.30
    BEV AP (Real)89.7987.4484.77
    BEV AP (Ours)82.8170.4465.68
    AOS AP (Real)90.6389.3488.36
    AOS AP (Ours)84.0269.6568.08
    下载: 导出CSV

    表  3  基于ShapeKITTI的三维目标检测模型(SECOND)平均精度 (%)

    Table  3  The average precision of 3D object detector (SECOND) based on ShapeKITTI dataset (%)

    评测项目简单模式中等模式困难模式
    3D AP (Real)87.4376.4869.10
    3D AP (Ours)59.6741.1837.67
    BEV AP (Real)89.5287.2883.89
    BEV AP (Ours)73.5553.3752.34
    AOS AP (Real)90.4989.4588.41
    AOS AP (Ours)75.0154.7254.13
    下载: 导出CSV

    表  4  基于闭环反馈的三维检测模型性能进化过程 (%)

    Table  4  The evolutionary process of 3D detection model based on the closed-loop feedback (%)

    序号AP (简单)AP (中等)AP (困难)反馈建议AP提升
    G010.6410.3210.53仅对少数目标能有效检测增加CAD模型种类
    G122.1416.4613.98目标尺寸估计欠佳改进CAD模型尺寸6.14 (G0 to G1)
    G232.8522.4019.52对稀疏目标检测效果欠佳降低前景目标密度5.94 (G1 to G2)
    G329.4326.8523.51目标高度估计效果欠佳调整前景目标高度4.45 (G2 to G3)
    G438.9929.7427.10对稀疏目标检测效果仍然欠佳进一步降低前景目标密度2.89 (G3 to G4)
    G539.2231.0430.302.30 (G4 to G5)
    下载: 导出CSV
  • [1] He K M, Zhang X Y, Ren S Q, Sun J. Residual learning for image recognition. In: Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, USA: IEEE, 2016. 770−778
    [2] Huang G, Liu Z, Van Der Maaten L, Weinberger K Q. Densely connected convolutional networks. In: Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu, USA: IEEE, 2017. 2261−2269
    [3] Ren S Q, He K M, Girshick R, Sun J. Faster R-CNN: Towards real-time object detection with region proposal networks. In: Proceedings of the 2015 Annual Conference on Neural Information Processing Systems. Montreal, Canada: 2015. 91−99
    [4] Zhu Z, Wu W, Zou W, Yan J J. End-to-end flow correlation tracking with spatial-temporal attention. In: Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, USA: IEEE, 2018. 548−557
    [5] Shen T Y, Wang J G, Gou C, Wang F-Y. Hierarchical fused model with deep learning and type-2 fuzzy learning for breast cancer diagnosis. IEEE Transactions on Fuzzy Systems, 2020, DOI: 10.119/TFUZZ.2020.3013681
    [6] Shen T Y, Gou C, Wang F-Y, He Z L, Chen W G. Learning from adversarial medical images for X-ray breast mass segmentation. Computer Methods and Programs in Biomedicine, 2019, 180: 105012 doi: 10.1016/j.cmpb.2019.105012
    [7] Chen X Z, Ma H M, Wan J, Li B, Xia T. Multi-view 3D object detection network for autonomous driving. In: Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu, USA: IEEE, 2017. 6526−6534
    [8] Wang C, Xu D F, Zhu Y K, Martín-Martín R, Lu C W, Li F F, Savarese S. Densefusion: 6D object pose estimation by iterative dense fusion. In: Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Long Beach, USA: IEEE, 2019. 3343−3352
    [9] Geiger A, Lenz P, Urtasun R. Are we ready for autonomous driving? The KITTI vision benchmark suite. In: Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition. Providence, USA: IEEE, 2012. 3354−3361
    [10] Sun P, Kretzschmar H, Dotiwalla X, Chouard A, Patnaik V, Tsui P, et al. Scalability in perception for autonomous driving: Waymo open dataset. In: Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, USA: IEEE, 2020. 2443−2451
    [11] Huang X Y, Wang P, Cheng X J, Zhou D F, Geng Q C, Yang R G. The apolloscape open dataset for autonomous driving and its application. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(10): 2702−2719 doi: 10.1109/TPAMI.2019.2926463
    [12] Wang Y, Chao W L, Garg D, Hariharan B, Campbell M, Weinberger K Q. Pseudo-LiDAR from visual depth estimation: Bridging the gap in 3D object detection for autonomous driving. In: Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Long Beach, USA: IEEE, 2019. 8437−8445
    [13] Qian R, Garg D, Wang Y, You Y R, Belongie S, Hariharan B, et al. End-to-end pseudo-LiDAR for image-based 3D object detection. In: Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, USA: IEEE, 2020. 5880−5889
    [14] Elmadawi K, Abdelrazek M, Elsobky M, Eraqi H M, Zahran M. End-to-end sensor modeling for LiDAR Point Cloud. In: Proceedings of the 2019 IEEE Intelligent Transportation Systems Conference (ITSC). IEEE, 2019. 1619−1624
    [15] Fang J, Zhou D F, Yan F L, Zhao T T, Zhang F H, Ma Y, et al. Augmented LiDAR simulator for autonomous driving. IEEE Robotics and Automation Letters, 2020, 5(2): 1931−1938 doi: 10.1109/LRA.2020.2969927
    [16] 王飞跃. 平行系统方法与复杂系统的管理和控制. 控制与决策, 2004, 19(5): 485−489, 514 doi: 10.3321/j.issn:1001-0920.2004.05.002

    Wang Fei-Yue. Parallel system methods for management and control of complex systems. Control and Decision, 2004, 19(5): 485−489, 514 doi: 10.3321/j.issn:1001-0920.2004.05.002
    [17] Wang F-Y. Parallel control and management for intelligent transportation systems: Concepts, architectures, and applications. IEEE Transactions on Intelligent Transportation Systems, 2010, 11(3): 630−638 doi: 10.1109/TITS.2010.2060218
    [18] 王飞跃. 平行控制: 数据驱动的计算控制方法. 自动化学报, 2013, 39(4): 293−302

    Wang Fei-Yue. Parallel control: A method for data-driven and computational control. Acta Automatica Sinica, 2013, 39(4): 293−302
    [19] Guo Y L, Wang H Y, Hu Q Y, Liu H, Liu L, Bennamoun M. Deep learning for 3D point clouds: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020 doi: 10.1109/TPAMI.2020.3005434
    [20] Su H, Maji S, Kalogerakis E, Learned-Miller E. Multi-view convolutional neural networks for 3D shape recognition. In: Proceedings of the 2015 IEEE International Conference on Computer Vision (ICCV). Santiago, Chile: IEEE, 2015. 945−953
    [21] Riegler G, Osman Ulusoy A, Geiger A. OctNet: Learning deep 3D representations at high resolutions. In: Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu, USA: IEEE, 2017. 6620−6629
    [22] Qi Charles R, Su H, Kaichun M, Guibas L J. Pointnet: Deep learning on point sets for 3D classification and segmentation. In: Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu, USA: IEEE, 2017. 77−85
    [23] Wu B C, Wan A, Yue X Y, Keutzer K. SqueezeSeg: Convolutional neural nets with recurrent CRF for real-time road-object segmentation from 3D LiDAR point cloud. In: Proceedings of the 2018 IEEE International Conference on Robotics and Automation (ICRA). Brisbane, Australia: IEEE, 2018. 1887−1893
    [24] Hou J, Dai A, Nießner M. 3D-SIS: 3D semantic instance segmentation of RGB-D scans. In: Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Long Beach, USA: IEEE, 2019. 4416−4425
    [25] Wang Z J, Lu F. VoxSegNet: Volumetric CNNs for semantic part segmentation of 3D shapes. IEEE Transactions on Visualization and Computer Graphics, 2020, 26(9): 2919−2930 doi: 10.1109/TVCG.2019.2896310
    [26] Lang A H, Vora S, Caesar H, Zhou L B, Yang J, Beijbom O. Pointpillars: Fast encoders for object detection from point clouds. In: Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Long Beach, USA: IEEE, 2019. 12689−12697
    [27] Tian Y L, Wang K F, Wang Y, Tian Y L, Wang Z L, Wang F-Y. Adaptive and azimuth-aware fusion network of multimodal local features for 3D object detection. Neurocomputing, 2020, 411: 32−44 doi: 10.1016/j.neucom.2020.05.086
    [28] Giancola S, Zarzar J, Ghanem B. Leveraging shape completion for 3D Siamese tracking. In: Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Long Beach, USA: IEEE, 2019. 1359−1368
    [29] Mueller M, Smith N, Ghanem B. Context-aware correlation filter tracking. In: Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu, USA: IEEE, 2017. 1387−1395
    [30] Caesar H, Bankiti V, Lang A H, Vora S, Liong V E, Xu Q, et al. nuScenes: A multimodal dataset for autonomous driving. In: Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, USA: IEEE, 2020. 11618−11628
    [31] Manivasagam S, Wang S, Wong K, Zeng W, Sazanovvich M, Tan S, Yang B, Ma W-C, Urtasun R. LiDARsim: Realistic LiDAR simulation by leveraging the real world. In: Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2020. 11167−11176
    [32] Prendinger H, Gajananan K, Zaki A B, Fares A, Molenaar R, Urbano D, et al. Tokyo virtual living lab: Designing smart cities based on the 3D Internet. IEEE Internet Computing, 2013, 17(6): 30−38 doi: 10.1109/MIC.2013.87
    [33] Ros G, Sellart L, Materzynska J, Vazquez D, López A M. The SYNTHIA dataset: A large collection of synthetic images for semantic segmentation of urban scenes. In: Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, USA: IEEE, 2016. 3234−3243
    [34] Gaidon A, Wang Q, Cabon Y, Vig E. VirtualWorlds as proxy for multi-object tracking analysis. In: Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, USA: IEEE, 2016. 4340−4349
    [35] Li X, Wang K F, Tian Y L, Yan L, Deng F, Wang F Y. The ParallelEye dataset: A large collection of virtual images for traffic vision research. IEEE Transactions on Intelligent Transportation Systems, 2019, 20(6): 2072−2084 doi: 10.1109/TITS.2018.2857566
    [36] Tian Y L, Li X, Wang K F, Wang F-Y. Training and testing object detectors with virtual images. IEEE/CAA Journal of Automatica Sinica, 2018, 5(2): 539−546 doi: 10.1109/JAS.2017.7510841
    [37] Johnson-Roberson M, Barto C, Mehta R, Sridhar S N, Rosaen K, Vasudevan R. Driving in the matrix: Can virtual worlds replace human-generated annotations for real world tasks? In: Proceedings of the 2017 IEEE International Conference on Robotics and Automation (ICRA). Singapore: IEEE, 2017. 746−753
    [38] Yue X Y, Wu B C, Seshia S A, Keutzer K, Sangiovanni-Vincentelli A L. A LiDAR point cloud generator: From a virtual world to autonomous driving. In: Proceedings of the 2018 ACM on International Conference on Multimedia Retrieval. Yokohama, Japan: ACM, 2018. 458−464
    [39] Li W, Pan C W, Zhang R, Ren J P, Ma Y X, Fang J, et al. AADS: Augmented autonomous driving simulation using data-driven algorithms. Science Robotics, 2019, 4(28): arXiv: 1901.07849v3 doi: 10.1126/scirobotics.aaw0863
    [40] 王坤峰, 苟超, 王飞跃. 平行视觉: 基于ACP的智能视觉计算方法. 自动化学报, 2016, 42(10): 1490−1500

    Wang Kun-Feng, Gou Chao, Wang Fei-Yue. Parallel vision: An ACP-based approach to intelligent vision computing. Acta Automatica Sinica, 2016, 42(10): 1490−1500
    [41] 王坤峰, 鲁越, 王雨桐, 熊子威, 王飞跃. 平行图像: 图像生成的一个新型理论框架. 模式识别与人工智能, 2017, 30(7): 577−587

    Wang Kun-Feng, Lu Yue, Wang Yu-Tong, Xiong Zi-Wei, Wang Fei-Yue. Parallel imaging: A new theoretical framework for image generation. Pattern Recognition and Artificial Intelligence, 2017, 30(7): 577−587
    [42] 李力, 林懿伦, 曹东璞, 郑南宁, 王飞跃. 平行学习 — 机器学习的一个新型理论框架. 自动化学报, 2017, 43(1): 1−8

    Li Li, Lin Yi-Lun, Cao Dong-Pu, Zheng Nan-Ning, Wang Fei-Yue. Parallel learning — A new framework for machine learning. Acta Automatica Sinica, 2017, 43(1): 1−8
    [43] Wang F-Y, Zheng N N, Cao D P, Martinez C M, Li L, Liu T. Parallel driving in CPSS: A unified approach for transport automation and vehicle intelligence. IEEE/CAA Journal of Automatica Sinica, 2017, 4(4): 577−587 doi: 10.1109/JAS.2017.7510598
    [44] 郭超, 鲁越, 林懿伦, 卓凡, 王飞跃. 平行艺术: 人机协作的艺术创作. 智能科学与技术学报, 2019, 1(4): 335−341

    Guo Chao, Lu Yue, Lin Yi-Lun, Zhuo Fan, Wang Fei-Yue. Parallel art: Artistic creation under human-machine collaboration. Chinese Journal of Intelligent Science and Technology, 2019, 1(4): 335−341
    [45] 吕宜生, 陈圆圆, 金峻臣, 李镇江, 叶佩军, 朱凤华. 平行交通: 虚实互动的智能交通管理与控制. 智能科学与技术学报, 2019, 1(1): 21−33 doi: 10.11959/j.issn.2096-6652.201908

    Lv Yi-Sheng, Chen Yuan-Yuan, Jin Jun-Chen, Li Zhen-Jiang, Ye Pei-Jun, Zhu Feng-Hua. Parallel transportation: Virtual-real interaction for intelligent traffic management and control. Chinese Journal of Intelligent Science and Technology, 2019, 1(1): 21−33 doi: 10.11959/j.issn.2096-6652.201908
    [46] 王飞跃. 平行传感与平行感知: 平行传感器的原理、框架与设计. 青岛智能产业技术研究院报告, 青岛, 中国, 2015.
    [47] Zhang H, Luo G Y, Tian Y L, Wang K F, He H B, Wang F-Y, et al. A virtual-real interaction approach to object instance segmentation in traffic scenes. IEEE Transactions on Intelligent Transportation Systems, 2020, DOI: 10.1109/TITS.2019.2961145.
    [48] Goodfellow I. NIPS 2016 tutorial: Generative adversarial networks. arXiv: 1701.00160, 2016.
    [49] Dosovitskiy A, Ros G, Codevilla F, López A M, Koltun V. CARLA: An open urban driving simulator. In: Proceedings of the 1st Annual Conference on Robot Learning. Mountain View, USA: PMLR, 2017. 1−16
    [50] Rong G D, Shin B H, Tabatabaee H, Lu Q, Lemke S, Možeiko M, et al. LGSVL simulator: A high fidelity simulator for autonomous driving. arXiv: 2005.03778, 2020.
    [51] Von Neumann-Cosel K, Dupuis M, Weiss C. Virtual test drive provision of a consistent tool-set for [D, H, S, V]-in-the-loop. In: Proceedings of the Driving Simulation Conference. Monaco, 2009.
    [52] You Y R, Wang Y, Chao W L, Garg D, Pleiss G, Hariharan B, et al. Pseudo-LiDAR++: Accurate depth for 3D object detection in autonomous driving. In: Proceedings of the 8th International Conference on Learning Representations. Addis Ababa, Ethiopia: OpenReview, 2020.
    [53] Chang A X, Funkhouser T, Guibas L, Hanrahan P, Huang Q X, Li Z M, et al. ShapeNet: An information-rich 3D model repository. arXiv: 1512.03012, 2015.
    [54] Yan Y, Mao Y X, Li B. SECOND: Sparsely embedded convolutional detection. Sensors, 2018, 18(10): 3337 doi: 10.3390/s18103337
  • 加载中
图(10) / 表(4)
计量
  • 文章访问数:  176
  • HTML全文浏览量:  197
  • PDF下载量:  125
  • 被引次数: 0
出版历程
  • 收稿日期:  2020-09-26
  • 网络出版日期:  2020-12-29
  • 刊出日期:  2020-12-29

目录

    /

    返回文章
    返回