3D semantic segmentation is one of the key tasks for autonomous driving system. Recently, deep learning models for 3D semantic segmentation task have been widely researched, but they usually require large amounts of training data. However, the present datasets for 3D semantic segmentation are lack of point-wise annotation, diversiform scenes and dynamic objects. In this paper1, we propose the SemanticPOSS dataset, which contains 2988 various and complicated LiDAR scans with large quantity of dynamic instances. The data is collected in Peking University and uses the same data format as SemanticKITTI. In addition, we evaluate several typical 3D semantic segmentation models on our SemanticPOSS dataset. Experimental results show that SemanticPOSS can help to improve the prediction accuracy of dynamic objects as people, car in some degree. SemanticPOSS will be published at www.poss.pku.edu.cn.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    SemanticPOSS: A Point Cloud Dataset with Large Quantity of Dynamic Instances


    Beteiligte:
    Pan, Yancheng (Autor:in) / Gao, Biao (Autor:in) / Mei, Jilin (Autor:in) / Geng, Sibo (Autor:in) / Li, Chengkun (Autor:in) / Zhao, Huijing (Autor:in)


    Erscheinungsdatum :

    19.10.2020


    Format / Umfang :

    3925789 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    ZPVehicles: a dataset of large vehicle 3D point cloud data

    Ye, Zhengzhou / Wang, Zihao / Chen, Xi et al. | IEEE | 2023


    VoxelScape: Large Scale Simulated 3D Point Cloud Dataset of Urban Traffic Environments

    Saleh, Khaled / Hossny, Mohammed / Abobakr, Ahmed et al. | IEEE | 2023


    Flexibility - the key to the automatic assembly of products in instances where quantity is low

    Dailami,F. / Stamp,R.J. / Bristol Polytechnic,GB | Kraftfahrwesen | 1989



    METHOD FOR ASSOCIATING POINT CLOUD DATA WITH RELEVANT DATA, PROGRAM, TRAINING DATASET, AND DEVICE

    KITAYAMA ATSUSHI / HYACINTH MARIO ARVIND ANAND / ANA-CRISTINA STAUDENMAIER | Europäisches Patentamt | 2024

    Freier Zugriff