English

Self-Supervised Visual Representation Learning from Hierarchical Grouping

Computer Vision and Pattern Recognition 2020-12-08 v1

Abstract

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.

Keywords

Cite

@article{arxiv.2012.03044,
  title  = {Self-Supervised Visual Representation Learning from Hierarchical Grouping},
  author = {Xiao Zhang and Michael Maire},
  journal= {arXiv preprint arXiv:2012.03044},
  year   = {2020}
}

Comments

Accepted by NeurIPS 2020

R2 v1 2026-06-23T20:45:09.216Z