LiDAR Point Clouds Semantic Segmentation in Autonomous Driving Based on Asymmetrical Convolution

被引:4
|
作者
Sun, Xiang [1 ]
Song, Shaojing [2 ]
Miao, Zhiqing [3 ]
Tang, Pan [4 ]
Ai, Luxia [5 ]
机构
[1] Shanghai Polytech Univ, Sch Intelligent Mfg & Control Engn, Shanghai 201209, Peoples R China
[2] Shanghai Polytech Univ, Sch Comp & Informat Engn, Shanghai 201209, Peoples R China
[3] East China Normal Univ, Sch Commun & Elect Engn, Shanghai 200062, Peoples R China
[4] Shanghai Univ, Sch Commun & Informat Engn, Shanghai 200444, Peoples R China
[5] Huazhong Univ Sci & Technol, Sch Artificial Intelligence & Automat, Wuhan 430074, Peoples R China
关键词
LiDAR point clouds; semantic segmentation; deep learning; asymmetric convolution; contextual feature enhancement;
D O I
10.3390/electronics12244926
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
LiDAR has become a vital sensor for autonomous driving scene understanding. To meet the accuracy and speed of LiDAR point clouds semantic segmentation, an efficient model ACPNet is proposed in this paper. In the feature extraction stage, the backbone is constructed with asymmetric convolutions, so the skeleton of the square convolution kernel is enhanced, which leads to greater robustness to target rotation. Moreover, a contextual feature enhancement module is designed to extract richer contextual features. During training, global scaling and global translation are performed to enrich the diversity of datasets. Compared with the baseline network PolarNet, the mIoU of ACPNet on the SemanticKITTI, SemanticPOSS and nuScenes datasets are improved by 5.1%, 1.6% and 2.9%, respectively. Meanwhile, the speed of ACPNet is 14 FPS, which basically meets the real-time requirements in autonomous driving scenarios. The experimental results show that ACPNet significantly improves the performance of LiDAR point cloud semantic segmentation.
引用
收藏
页数:13
相关论文
共 50 条
  • [1] SalsaNet: Fast Road and Vehicle Segmentation in LiDAR Point Clouds for Autonomous Driving
    Aksoy, Eren Erdal
    Baci, Saimir
    Cavdar, Selcuk
    [J]. 2020 IEEE INTELLIGENT VEHICLES SYMPOSIUM (IV), 2020, : 926 - 932
  • [2] Real-Time LiDAR Point Cloud Semantic Segmentation for Autonomous Driving
    Xie, Xing
    Bai, Lin
    Huang, Xinming
    [J]. ELECTRONICS, 2022, 11 (01)
  • [3] Autonomous driving semantic segmentation with convolution neural networks
    Wang, Zhong-Yu
    Ni, Xian-Yang
    Shang, Zhen-Dong
    [J]. Guangxue Jingmi Gongcheng/Optics and Precision Engineering, 2019, 27 (11): : 2429 - 2438
  • [4] A Cylindrical Convolution Network for Dense Top-View Semantic Segmentation with LiDAR Point Clouds
    Lu, Jiacheng
    Gu, Shuo
    Xu, Cheng-Zhong
    Kong, Hui
    [J]. COMPUTER VISION - ACCV 2022, PT VII, 2023, 13847 : 344 - 360
  • [5] LIGHTNINGNET : FAST AND ACCURATE SEMANTIC SEGMENTATION FOR AUTONOMOUS DRIVING BASED ON 3D LIDAR POINT CLOUD
    Yang, Kaihong
    Bi, Sheng
    Dong, Min
    [J]. 2020 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO (ICME), 2020,
  • [6] Semantic Segmentation Method for Road Intersection Point Clouds Based on Lightweight LiDAR
    Ren, Xiaole
    Yu, Bin
    Wang, Yuchen
    [J]. APPLIED SCIENCES-BASEL, 2024, 14 (11):
  • [7] Scan-based Semantic Segmentation of LiDAR Point Clouds: An Experimental Study
    Triess, Larissa T.
    Peter, David
    Rist, Christoph B.
    Zoellner, J. Marius
    [J]. 2020 IEEE INTELLIGENT VEHICLES SYMPOSIUM (IV), 2020, : 1116 - 1121
  • [8] Semantic Segmentation and Roof Reconstruction of Urban Buildings Based on LiDAR Point Clouds
    Sun, Xiaokai
    Guo, Baoyun
    Li, Cailin
    Sun, Na
    Wang, Yue
    Yao, Yukai
    [J]. ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION, 2024, 13 (01)
  • [9] Semantic Segmentation of Spectral LiDAR Point Clouds Based on Neural Architecture Search
    Zhang, Qi
    Peng, Yuanxi
    Zhang, Zhiwen
    Li, Teng
    [J]. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2023, 61
  • [10] Exploring Adversarial Robustness of LiDAR Semantic Segmentation in Autonomous Driving
    Mahima, K. T. Yasas
    Perera, Asanka
    Anavatti, Sreenatha
    Garratt, Matt
    [J]. SENSORS, 2023, 23 (23)