Skip to ContentSkip to Navigation
About us Latest news Events PhD ceremonies

Automating the detection of strong gravitational lenses in large-scale surveys using deep learning

PhD ceremony:B.C. Nagam
When:February 10, 2025
Start:12:45
Supervisors:prof. dr. L.V.E. (Léon) Koopmans, prof. dr. E.A. (Edwin A.) Valentijn
Co-supervisor:dr. G.A. (Gijs) Verdoes Kleijn
Where:Academy building RUG
Faculty:Science and Engineering
Automating the detection of strong gravitational lenses in
large-scale surveys using deep learning

A gravitational lens is an effect that occurs due to the curvature of space-time, when a massive object in the forefront, such a galaxy or a cluster of galaxies, bends and magnifies the light of a source behind it. This may result in multiple images, curves, or even complete Einstein-rings of the background object appearing around the forefront lens mass. 

In his thesis, Bharath Chowdhary Nagam presents advancements in automated strong gravitational lens detection using novel CNN-based architectures and techniques, focusing on reducing false positives - a major challenge in large-scale surveys. The DenseNet-121 ensemble outperforms previous ResNet models, achieving higher true-positive rates with fewer parameters. Nagam introduces a novel Information Content (IC) metric for ranking lens candidates, improving classification performance. Integration with a U-Net segmentation algorithm (U-Denselens) enhances detection by filtering false positives based on segmentation scores, achieving substantial reductions in false positives while retaining true lenses, as demonstrated on KiDS and Euclid datasets.

To further enhance training, Denoising Diffusion GANs (DDGANs) generate large datasets of mock lenses. A balanced combination of real and synthetic data improves CNN's performance, addressing data scarcity in upcoming surveys. The U-DenseLens model applied to Euclid Early Release Observations (ERO) achieves lower false-positive rates and high detection efficiency, identifying 46 lens candidates across 16 fields, with contamination rates significantly lower than for KiDS data. Scaling projections suggest discovering over 5,500 strong lenses across the Euclid survey area.

These results underscore the potential of advanced CNN architectures, generative models, and segmentation algorithms to automate lens detection. By integrating multiple metrics, this framework offers a promising pathway to optimizing lens searches in large-scale astronomical datasets.