Clustering via Self-Supervised Diffusion

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Uziel, Roy, Chelly, Irit, Freifeld, Oren, Pakman, Ari
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913965254115328
author Uziel, Roy
Chelly, Irit
Freifeld, Oren
Pakman, Ari
author_facet Uziel, Roy
Chelly, Irit
Freifeld, Oren
Pakman, Ari
contents Diffusion models, widely recognized for their success in generative tasks, have not yet been applied to clustering. We introduce Clustering via Diffusion (CLUDI), a self-supervised framework that combines the generative power of diffusion models with pre-trained Vision Transformer features to achieve robust and accurate clustering. CLUDI is trained via a teacher-student paradigm: the teacher uses stochastic diffusion-based sampling to produce diverse cluster assignments, which the student refines into stable predictions. This stochasticity acts as a novel data augmentation strategy, enabling CLUDI to uncover intricate structures in high-dimensional data. Extensive evaluations on challenging datasets demonstrate that CLUDI achieves state-of-the-art performance in unsupervised classification, setting new benchmarks in clustering robustness and adaptability to complex data distributions. Our code is available at https://github.com/BGU-CS-VIL/CLUDI.
format Preprint
id arxiv_https___arxiv_org_abs_2507_04283
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Clustering via Self-Supervised Diffusion
Uziel, Roy
Chelly, Irit
Freifeld, Oren
Pakman, Ari
Artificial Intelligence
Computer Vision and Pattern Recognition
Diffusion models, widely recognized for their success in generative tasks, have not yet been applied to clustering. We introduce Clustering via Diffusion (CLUDI), a self-supervised framework that combines the generative power of diffusion models with pre-trained Vision Transformer features to achieve robust and accurate clustering. CLUDI is trained via a teacher-student paradigm: the teacher uses stochastic diffusion-based sampling to produce diverse cluster assignments, which the student refines into stable predictions. This stochasticity acts as a novel data augmentation strategy, enabling CLUDI to uncover intricate structures in high-dimensional data. Extensive evaluations on challenging datasets demonstrate that CLUDI achieves state-of-the-art performance in unsupervised classification, setting new benchmarks in clustering robustness and adaptability to complex data distributions. Our code is available at https://github.com/BGU-CS-VIL/CLUDI.
title Clustering via Self-Supervised Diffusion
topic Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2507.04283