We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

astro-ph

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computer Vision and Pattern Recognition

Title: Towards Galaxy Foundation Models with Hybrid Contrastive Learning

Abstract: New astronomical tasks are often related to earlier tasks for which labels have already been collected. We adapt the contrastive framework BYOL to leverage those labels as a pretraining task while also enforcing augmentation invariance. For large-scale pretraining, we introduce GZ-Evo v0.1, a set of 96.5M volunteer responses for 552k galaxy images plus a further 1.34M comparable unlabelled galaxies. Most of the 206 GZ-Evo answers are unknown for any given galaxy, and so our pretraining task uses a Dirichlet loss that naturally handles unknown answers. GZ-Evo pretraining, with or without hybrid learning, improves on direct training even with plentiful downstream labels (+4% accuracy with 44k labels). Our hybrid pretraining/contrastive method further improves downstream accuracy vs. pretraining or contrastive learning, especially in the low-label transfer regime (+6% accuracy with 750 labels).
Comments: Accepted at the ICML 2022 Workshop on Machine Learning for Astrophysics. Data: www.github.com/mwalmsley/pytorch-galaxy-datasets. Please reach out to share your labelled data - all contributions will be credited in future work
Subjects: Computer Vision and Pattern Recognition (cs.CV); Astrophysics of Galaxies (astro-ph.GA)
Cite as: arXiv:2206.11927 [cs.CV]
  (or arXiv:2206.11927v1 [cs.CV] for this version)

Submission history

From: Mike Walmsley [view email]
[v1] Thu, 23 Jun 2022 18:07:27 GMT (714kb,D)

Link back to: arXiv, form interface, contact.