Self-Supervised Visual Representation Learning from Hierarchical Grouping

被引:0
|
作者
Zhang, Xiao [1 ]
Maire, Michael [1 ]
机构
[1] Univ Chicago, Chicago, IL 60637 USA
来源
ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 33, NEURIPS 2020 | 2020年 / 33卷
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We create a framework for bootstrapping visual representation learning from a primitive visual grouping capability. We operationalize grouping via a contour detector that partitions an image into regions, followed by merging of those regions into a tree hierarchy. A small supervised dataset suffices for training this grouping primitive. Across a large unlabeled dataset, we apply this learned primitive to automatically predict hierarchical region structure. These predictions serve as guidance for self-supervised contrastive feature learning: we task a deep network with producing per-pixel embeddings whose pairwise distances respect the region hierarchy. Experiments demonstrate that our approach can serve as state-of-the-art generic pre-training, benefiting downstream tasks. We additionally explore applications to semantic region search and video-based object instance tracking.
引用
收藏
页数:12
相关论文
共 50 条
  • [31] Semantics-Consistent Feature Search for Self-Supervised Visual Representation Learning
    Song, Kaiyou
    Zhang, Shan
    Luo, Zimeng
    Wang, Tong
    Xie, Jin
    2023 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2023), 2023, : 16053 - 16062
  • [32] Feature selection and cascade dimensionality reduction for self-supervised visual representation learning
    Qu, Peixin
    Jin, Songlin
    Tian, Yongqin
    Zhou, Ling
    Zheng, Ying
    Zhang, Weidong
    Xu, Yibo
    Pan, Xipeng
    Zhao, Wenyi
    COMPUTERS & ELECTRICAL ENGINEERING, 2023, 106
  • [33] Self-Supervised Representation Learning using Visual Field Expansion on Digital Pathology
    Boyd, Joseph
    Liashuha, Mykola
    Deutsch, Eric
    Paragios, Nikos
    Christodoulidis, Stergios
    Vakalopoulou, Maria
    2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION WORKSHOPS (ICCVW 2021), 2021, : 639 - 647
  • [34] Revitalizing CNN Attentions via Transformers in Self-Supervised Visual Representation Learning
    Ge, Chongjian
    Liang, Youwei
    Song, Yibing
    Jiao, Jianbo
    Wang, Jue
    Luo, Ping
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 34 (NEURIPS 2021), 2021, 34
  • [35] Self-Supervised Audio-Visual Representation Learning for in-the-wild Videos
    Feng, Zishun
    Tu, Ming
    Xia, Rui
    Wang, Yuxuan
    Krishnamurthy, Ashok
    2020 IEEE INTERNATIONAL CONFERENCE ON BIG DATA (BIG DATA), 2020, : 5671 - 5672
  • [36] solo-learn: A Library of Self-supervised Methods for Visual Representation Learning
    Turrisi da Costa, Victor G.
    Fini, Enrico
    Nabi, Moin
    Sebe, Nicu
    Ricci, Elisa
    JOURNAL OF MACHINE LEARNING RESEARCH, 2022, 23 : 1 - 6
  • [37] Self-Supervised Speech Representation Learning: A Review
    Mohamed, Abdelrahman
    Lee, Hung-yi
    Borgholt, Lasse
    Havtorn, Jakob D.
    Edin, Joakim
    Igel, Christian
    Kirchhoff, Katrin
    Li, Shang-Wen
    Livescu, Karen
    Maaloe, Lars
    Sainath, Tara N.
    Watanabe, Shinji
    IEEE JOURNAL OF SELECTED TOPICS IN SIGNAL PROCESSING, 2022, 16 (06) : 1179 - 1210
  • [38] Distilling Localization for Self-Supervised Representation Learning
    Zhao, Nanxuan
    Wu, Zhirong
    Lau, Rynson W. H.
    Lin, Stephen
    THIRTY-FIFTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THIRTY-THIRD CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE AND THE ELEVENTH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2021, 35 : 10990 - 10998
  • [39] Self-Supervised Relational Reasoning for Representation Learning
    Patacchiola, Massimiliano
    Storkey, Amos
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 33, NEURIPS 2020, 2020, 33
  • [40] Self-Supervised Learning for Specified Latent Representation
    Liu, Chicheng
    Song, Libin
    Zhang, Jiwen
    Chen, Ken
    Xu, Jing
    IEEE TRANSACTIONS ON FUZZY SYSTEMS, 2020, 28 (01) : 47 - 59