Work

Group-invariant Coresets for Data-efficient Active Learning

Year: 2026

Type: preprint

Abstract: Active learning reduces labeling cost by querying the most informative unlabeled samples, but standard coreset methods ignore known data symmetries and can waste budget on transformed versions of the same instance. We propose GRINCO, a group-invariant coreset framework that performs acquisition in the quotient space induced by a transformation group, so that selection operates on orbits rather than raw samples. The method uses either canonical representatives or learned orbit-separating invariant embeddings to define practical quotient metrics, and combines quotient-space k-center selection with invariant training through an orbit-averaged loss. We further derive a generalization bound that relates excess orbit-averaged risk to quotient-space coverage, label uncertainty, and intra-orbit variability. Experiments on synthetic scale-invariant data and image benchmarks with rotation-induced redundancy show that GRINCO improves orbit coverage and achieves stronger label efficiency than conventional coreset baselines, especially when group-induced redundancy is substantial.

Source: arXiv (Cornell University)

DOI: 10.48550/arxiv.2607.01089

Authors: Ayres, L. C., Bermudez, J. C. M., de Almeida, S. J. M., Borsoi, R. A.

Institution: Universidade Federal de Santa Catarina

Topic: Machine Learning and Algorithms

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

Funders: Agence Nationale de la Recherche, Coordenação de Aperfeiçoamento de Pessoal de Nível Superior

Awards: Scalable online change-points and events detection for dynamic structured data, G29316538, LEarning and adapting geNeraTIve models to soLve blind inverse probLEms, Collaborative Research: CISE-ANR: CIF: Small: Learning from large datasets: Application to multi-subject fMRI analysis

OpenAlex ID: W7167114402

Cites:

Cited by:

Related to: