LenNet: Direct Detection and Localization of Strong Gravitational Lenses in Wide-Field Sky Survey Images
Abstract
Strong gravitational lenses are invaluable tools for addressing fundamental questions in astrophysics, from the nature of dark matter to the expansion of the universe. While current sky surveys have successfully identified thousands of lens candidates, the search methods employed face a critical challenge. The conventional approach relies on a "crop-and-classify" strategy, where small images are first cut out around billions of potential host galaxies before being individually classified. This process creates a significant computational and storage bottleneck that is unsustainable for future large-scale surveys. To overcome this limitation, we propose LenNet, an object detection model that identifies lenses directly within large, original survey images. Our method completely bypasses the inefficient cropping step by framing the problem as a direct detection and localization task. We initially train LenNet on simulated data to learn the complex features of gravitational lenses and then use transfer learning to fine-tune the model on a limited set of real, labeled examples from the Kilo-Degree Survey (KiDS). Our experiments show that LenNet performs remarkably well on real survey data, validating its potential as a highly efficient and scalable solution for lens discovery in massive astronomical surveys.
Versions
Related
- Polar coordinate transformations for machine-learning-based dark matter subhalo detection in strong gravitational lenses
- Hints of self-interacting dark matter from strong gravitational lensing
- LUMA: A CNN for Strong Gravitational Lens Searches in Astronomical Imaging
- On the Angular Localization of Gravitational-wave Signals by Pulsar Timing Arrays
- Gravitational waves from strong first-order phase transitions in an expanding Universe
BibTeX
@misc{liu2026lennet,
title = {{LenNet: Direct Detection and Localization of Strong Gravitational Lenses in Wide-Field Sky Survey Images}},
author = {Pufan Liu and Hui Li and Ziqi Li and Xiaoyue Cao and Rui Li and Hao Su and Ran Li and Nicola R. Napolitano and Léon V. E. Koopmans and Valerio Busillo and Crescenzo Tortora and Liang Gao},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.21661},
eprint = {2609.21661},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.21661},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - LenNet: Direct Detection and Localization of Strong Gravitational Lenses in Wide-Field Sky Survey Images
AU - Pufan Liu
AU - Hui Li
AU - Ziqi Li
AU - Xiaoyue Cao
AU - Rui Li
AU - Hao Su
AU - Ran Li
AU - Nicola R. Napolitano
AU - Léon V. E. Koopmans
AU - Valerio Busillo
AU - Crescenzo Tortora
AU - Liang Gao
PY - 2026
DA - 2026/09/18
PB - arXiv
DO - 10.48550/arxiv.2609.21661
UR - https://arxiv.org/abs/2609.21661
AB - Strong gravitational lenses are invaluable tools for addressing fundamental questions in astrophysics, from the nature of dark matter to the expansion of the universe. While current sky surveys have successfully identified thousands of lens candidates, the search methods employed face a critical challenge. The conventional approach relies on a "crop-and-classify" strategy, where small images are first cut out around billions of potential host galaxies before being individually classified. This process creates a significant computational and storage bottleneck that is unsustainable for future large-scale surveys. To overcome this limitation, we propose LenNet, an object detection model that identifies lenses directly within large, original survey images. Our method completely bypasses the inefficient cropping step by framing the problem as a direct detection and localization task. We initially train LenNet on simulated data to learn the complex features of gravitational lenses and then use transfer learning to fine-tune the model on a limited set of real, labeled examples from the Kilo-Degree Survey (KiDS). Our experiments show that LenNet performs remarkably well on real survey data, validating its potential as a highly efficient and scalable solution for lens discovery in massive astronomical surveys.
ER -