Publication Type
Conference Proceeding Article
Version
acceptedVersion
Publication Date
12-2021
Abstract
A good visual representation is an inference map from observations (images) to features (vectors) that faithfully reflects the hidden modularized generative factors (semantics). In this paper, we formulate the notion of “good” representation from a group-theoretic view using Higgins’ definition of disentangled representation [38], and show that existing Self-Supervised Learning (SSL) only disentangles simple augmentation features such as rotation and colorization, thus unable to modularize the remaining semantics. To break the limitation, we propose an iterative SSL algorithm: Iterative Partition-based Invariant Risk Minimization (IP-IRM), which successfully grounds the abstract semantics and the group acting on them into concrete contrastive learning. At each iteration, IP-IRM first partitions the training samples into two subsets that correspond to an entangled group element. Then, it minimizes a subset-invariant contrastive loss, where the invariance guarantees to disentangle the group element. We prove that IP-IRM converges to a fully disentangled representation and show its effectiveness on various benchmarks.
Discipline
Databases and Information Systems | Graphics and Human Computer Interfaces
Research Areas
Intelligent Systems and Optimization
Publication
Proceedings of the 35th Conference on Neural Information Processing Systems, Sydney, Australia, 2021 December 6-14
First Page
1
Last Page
8
City or Country
Virtual Conference
Citation
WANG, Tan; YUE, Zhongqi; HUANG, Jianqiang; SUN, Qianru; and ZHANG, Hanwang.
Self-supervised learning disentangled group representation as feature. (2021). Proceedings of the 35th Conference on Neural Information Processing Systems, Sydney, Australia, 2021 December 6-14. 1-8.
Available at: https://ink.library.smu.edu.sg/sis_research/6227
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Included in
Databases and Information Systems Commons, Graphics and Human Computer Interfaces Commons