LenSiam: Self-Supervised Learning on Strong Gravitational Lens Images. (arXiv:2311.10100v1 [astro-ph.IM])
<a href="http://arxiv.org/find/astro-ph/1/au:+Chang_P/0/1/0/all/0/1">Po-Wen Chang</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Huang_K/0/1/0/all/0/1">Kuan-Wei Huang</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Fagin_J/0/1/0/all/0/1">Joshua Fagin</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Chan_J/0/1/0/all/0/1">James Hung-Hsu Chan</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Lin_J/0/1/0/all/0/1">Joshua Yao-Yu Lin</a>
Self-supervised learning has been known for learning good representations
from data without the need for annotated labels. We explore the simple siamese
(SimSiam) architecture for representation learning on strong gravitational lens
images. Commonly used image augmentations tend to change lens properties; for
example, zoom-in would affect the Einstein radius. To create image pairs
representing the same underlying lens model, we introduce a lens augmentation
method to preserve lens properties by fixing the lens model while varying the
source galaxies. Our research demonstrates this lens augmentation works well
with SimSiam for learning the lens image representation without labels, so we
name it LenSiam. We also show that a pre-trained LenSiam model can benefit
downstream tasks. We open-source our code and datasets at
https://github.com/kuanweih/LenSiam .
Self-supervised learning has been known for learning good representations
from data without the need for annotated labels. We explore the simple siamese
(SimSiam) architecture for representation learning on strong gravitational lens
images. Commonly used image augmentations tend to change lens properties; for
example, zoom-in would affect the Einstein radius. To create image pairs
representing the same underlying lens model, we introduce a lens augmentation
method to preserve lens properties by fixing the lens model while varying the
source galaxies. Our research demonstrates this lens augmentation works well
with SimSiam for learning the lens image representation without labels, so we
name it LenSiam. We also show that a pre-trained LenSiam model can benefit
downstream tasks. We open-source our code and datasets at
https://github.com/kuanweih/LenSiam .
http://arxiv.org/icons/sfx.gif
