Masked Autoencoder Pretraining on Strong-Lensing Images for Joint Dark-Matter Model Classification and Super-Resolution
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866908697009061888 |
|---|---|
| author | Prasha, Achmad Ardani Rachmadi, Clavino Ourizqi Syahlan, Muhamad Fauzan Ibnu Anugerah, Naufal Rahfi Raditya, Nanda Garin Amelia, Putri Mutiara, Sabrina Laila Ramadhan, Hilman Syachr |
| author_facet | Prasha, Achmad Ardani Rachmadi, Clavino Ourizqi Syahlan, Muhamad Fauzan Ibnu Anugerah, Naufal Rahfi Raditya, Nanda Garin Amelia, Putri Mutiara, Sabrina Laila Ramadhan, Hilman Syachr |
| contents | Strong gravitational lensing can reveal the influence of dark-matter substructure in galaxies, but analyzing these effects from noisy, low-resolution images poses a significant challenge. In this work, we propose a masked autoencoder (MAE) pretraining strategy on simulated strong-lensing images from the DeepLense ML4SCI benchmark to learn generalizable representations for two downstream tasks: (i) classifying the underlying dark matter model (cold dark matter, axion-like, or no substructure) and (ii) enhancing low-resolution lensed images via super-resolution. We pretrain a Vision Transformer encoder using a masked image modeling objective, then fine-tune the encoder separately for each task. Our results show that MAE pretraining, when combined with appropriate mask ratio tuning, yields a shared encoder that matches or exceeds a ViT trained from scratch. Specifically, at a 90% mask ratio, the fine-tuned classifier achieves macro AUC of 0.968 and accuracy of 88.65%, compared to the scratch baseline (AUC 0.957, accuracy 82.46%). For super-resolution (16x16 to 64x64), the MAE-pretrained model reconstructs images with PSNR ~33 dB and SSIM 0.961, modestly improving over scratch training. We ablate the MAE mask ratio, revealing a consistent trade-off: higher mask ratios improve classification but slightly degrade reconstruction fidelity. Our findings demonstrate that MAE pretraining on physics-rich simulations provides a flexible, reusable encoder for multiple strong-lensing analysis tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_06642 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Masked Autoencoder Pretraining on Strong-Lensing Images for Joint Dark-Matter Model Classification and Super-Resolution Prasha, Achmad Ardani Rachmadi, Clavino Ourizqi Syahlan, Muhamad Fauzan Ibnu Anugerah, Naufal Rahfi Raditya, Nanda Garin Amelia, Putri Mutiara, Sabrina Laila Ramadhan, Hilman Syachr Computer Vision and Pattern Recognition Cosmology and Nongalactic Astrophysics Instrumentation and Methods for Astrophysics Artificial Intelligence Machine Learning Strong gravitational lensing can reveal the influence of dark-matter substructure in galaxies, but analyzing these effects from noisy, low-resolution images poses a significant challenge. In this work, we propose a masked autoencoder (MAE) pretraining strategy on simulated strong-lensing images from the DeepLense ML4SCI benchmark to learn generalizable representations for two downstream tasks: (i) classifying the underlying dark matter model (cold dark matter, axion-like, or no substructure) and (ii) enhancing low-resolution lensed images via super-resolution. We pretrain a Vision Transformer encoder using a masked image modeling objective, then fine-tune the encoder separately for each task. Our results show that MAE pretraining, when combined with appropriate mask ratio tuning, yields a shared encoder that matches or exceeds a ViT trained from scratch. Specifically, at a 90% mask ratio, the fine-tuned classifier achieves macro AUC of 0.968 and accuracy of 88.65%, compared to the scratch baseline (AUC 0.957, accuracy 82.46%). For super-resolution (16x16 to 64x64), the MAE-pretrained model reconstructs images with PSNR ~33 dB and SSIM 0.961, modestly improving over scratch training. We ablate the MAE mask ratio, revealing a consistent trade-off: higher mask ratios improve classification but slightly degrade reconstruction fidelity. Our findings demonstrate that MAE pretraining on physics-rich simulations provides a flexible, reusable encoder for multiple strong-lensing analysis tasks. |
| title | Masked Autoencoder Pretraining on Strong-Lensing Images for Joint Dark-Matter Model Classification and Super-Resolution |
| topic | Computer Vision and Pattern Recognition Cosmology and Nongalactic Astrophysics Instrumentation and Methods for Astrophysics Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2512.06642 |