T-REGS: Minimum Spanning Tree Regularization for Self-Supervised Learning

Julie Mordacq, David Loiseaux, Vicky Kalogeiton, Steve Oudot

Keywords: self-supervised learning, minimum spanning tree, dimensional collapse, dimension estimation, topological data analysis

TL;DR: We introduce a minimum-spanning-tree-based regularization for self-supervised learning that provably improves representation uniformity and mitigates dimensional collapse.

Abstract:

Self-supervised learning (SSL) has emerged as a powerful paradigm for learning representations without labeled data, often by enforcing invariance to input transformations such as rotations or blurring.
Recent studies have highlighted two pivotal properties for effective representations: (i) avoiding dimensional collapse-where the learned features occupy only a low-dimensional subspace, and (ii) enhancing uniformity of the induced distribution.
In this work, we introduce T-REGS, a simple regularization framework for SSL based on the length of the Minimum Spanning Tree (MST) over the learned representation.
We provide theoretical analysis demonstrating that T-REGS simultaneously mitigates dimensional collapse and promotes distribution uniformity on arbitrary compact Riemannian manifolds.
Several experiments on synthetic data and on classical SSL benchmarks validate the effectiveness of our approach at enhancing representation quality.

Primary Area: General machine learning (supervised, unsupervised, online, active, etc.)

Submission Number: 28174