Complete-to-Partial 4D Distillation for Self-Supervised Point Cloud Sequence Representation Learning

被引:2
|
作者
Zhang, Zhuoyang [1 ]
Dong, Yuhao [1 ]
Liu, Yunze [1 ,3 ]
Yi, Li [1 ,2 ,3 ]
机构
[1] Tsinghua Univ, IIIS, Beijing, Peoples R China
[2] Shanghai Artificial Intelligence Lab, Shanghai, Peoples R China
[3] Shanghai Qi Zhi Inst, Shanghai, Peoples R China
关键词
D O I
10.1109/CVPR52729.2023.01694
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Recent work on 4D point cloud sequences has attracted a lot of attention. However, obtaining exhaustively labeled 4D datasets is often very expensive and laborious, so it is especially important to investigate how to utilize raw unlabeled data. However, most existing self-supervised point cloud representation learning methods only consider geometry from a static snapshot omitting the fact that sequential observations of dynamic scenes could reveal more comprehensive geometric details. To overcome such issues, this paper proposes a new 4D self-supervised pre-training method called Complete-to-Partial 4D Distillation. Our key idea is to formulate 4D self-supervised representation learning as a teacher-student knowledge distillation framework and let the student learn useful 4D representations with the guidance of the teacher. Experiments show that this approach significantly outperforms previous pre-training approaches on a wide range of 4D point cloud sequence understanding tasks. Code is available at: https://github.com/dongyh20/C2P.
引用
收藏
页码:17661 / 17670
页数:10
相关论文
共 50 条
  • [1] Implicit Autoencoder for Point-Cloud Self-Supervised Representation Learning
    Yan, Siming
    Yang, Zhenpei
    Li, Haoxiang
    Song, Chen
    Guan, Li
    Kang, Hao
    Hua, Gang
    Huang, Qixing
    2023 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2023), 2023, : 14484 - 14496
  • [2] Self-Supervised Learning of Part Mobility from Point Cloud Sequence
    Shi, Yahao
    Cao, Xinyu
    Zhou, Bin
    COMPUTER GRAPHICS FORUM, 2021, 40 (06) : 104 - 116
  • [3] Generative Variational-Contrastive Learning for Self-Supervised Point Cloud Representation
    Wang, Bohua
    Tian, Zhiqiang
    Ye, Aixue
    Wen, Feng
    Du, Shaoyi
    Gao, Yue
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2024, 46 (09) : 6154 - 6166
  • [4] Self-Supervised Point Cloud Representation Learning via Separating Mixed Shapes
    Sun, Chao
    Zheng, Zhedong
    Wang, Xiaohan
    Xu, Mingliang
    Yang, Yi
    IEEE TRANSACTIONS ON MULTIMEDIA, 2023, 25 : 6207 - 6218
  • [5] A COMPREHENSIVE STUDY ON SELF-SUPERVISED DISTILLATION FOR SPEAKER REPRESENTATION LEARNING
    Chen, Zhengyang
    Qian, Yao
    Han, Bing
    Qian, Yanmin
    Zeng, Michael
    2022 IEEE SPOKEN LANGUAGE TECHNOLOGY WORKSHOP, SLT, 2022, : 599 - 604
  • [6] Self-supervised 4D Spatio-temporal Feature Learning via Order Prediction of Sequential Point Cloud Clips
    Wang, Haiyan
    Yang, Liang
    Rong, Xuejian
    Feng, Jinglun
    Tian, Yingli
    2021 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION WACV 2021, 2021, : 3761 - 3770
  • [7] Self-supervised learning for point cloud data: A survey
    Zeng, Changyu
    Wang, Wei
    Nguyen, Anh
    Xiao, Jimin
    Yue, Yutao
    EXPERT SYSTEMS WITH APPLICATIONS, 2024, 237
  • [8] Masked Autoencoders for Point Cloud Self-supervised Learning
    Pang, Yatian
    Wang, Wenxiao
    Tay, Francis E. H.
    Liu, Wei
    Tian, Yonghong
    Yuan, Li
    COMPUTER VISION - ECCV 2022, PT II, 2022, 13662 : 604 - 621
  • [9] SegContrast: 3D Point Cloud Feature Representation Learning Through Self-Supervised Segment Discrimination
    Nunes, Lucas
    Marcuzzi, Rodrigo
    Chen, Xieyuanli
    Behley, Jens
    Stachniss, Cyrill
    IEEE ROBOTICS AND AUTOMATION LETTERS, 2022, 7 (02) : 2116 - 2123
  • [10] Self-Supervised 3D Action Representation Learning With Skeleton Cloud Colorization
    Yang, Siyuan
    Liu, Jun
    Lu, Shijian
    Hwa, Er Meng
    Hu, Yongjian
    Kot, Alex C.
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2024, 46 (01) : 509 - 524