English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

While supervised neural networks have become state of the art for identifyingthe rare strong gravitational lenses from large imaging data sets, theirselection remains significantly affected by the large number and diversity ofnonlens contaminants. This work evaluates and compares systematically theperformance of neural networks in order to move towards a rapid selection ofgalaxy-scale strong lenses with minimal human input in the era of deep,wide-scale surveys. We used multiband images from PDR2 of the HSC Wide surveyto build test sets mimicking an actual classification experiment, with 189strong lenses previously found over the HSC footprint and 70,910 nonlensgalaxies in COSMOS. Multiple networks were trained on different sets ofrealistic strong-lens simulations and nonlens galaxies, with variousarchitectures and data pre-processing. The overall performances strongly dependon the construction of the ground-truth training data and they typically, butnot systematically, improve using our baseline residual network architecture.Improvements are found when applying random shifts to the image centroids andsquare root stretches to the pixel values, adding z band, or using randomviewpoints of the original images, but not when adding difference images tosubtract emission from the central galaxy. The most significant gain isobtained with committees of networks trained on different data sets, andshowing a moderate overlap between populations of false positives.Nearly-perfect invariance to image quality can be achieved by training networkseither with large number of bands, or jointly with the PSF and science frames.Overall, we show the possibility to reach a TPR0 as high as 60% for the testsets under consideration, which opens promising perspectives for pure selectionof strong lenses without human input using the Rubin Observatory and otherforthcoming ground-based surveys.

HOLISMOKES -- XI. Evaluation of supervised neural networks for strong-lens searches in ground-based imaging surveys