Image-Based Multi-Survey Classification of Light Curves with a Pre-Trained Vision Transformer

2025/07/15 by Moreno-Cartagena, Daniel, Cabrera-Vives, Guillermo, Arancibia, Alejandra M. Muñoz +10
#Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #FOS: Physical sciences #Instrumentation and Methods for Astrophysics (astro-ph.IM)

paper · doi:10.48550/arxiv.2507.11711

Abstract

We explore the use of Swin Transformer V2, a pre-trained vision Transformer, for photometric classification in a multi-survey setting by leveraging light curves from the Zwicky Transient Facility (ZTF) and the Asteroid Terrestrial-impact Last Alert System (ATLAS). We evaluate different strategies for integrating data from these surveys and find that a multi-survey architecture which processes them jointly achieves the best performance. These results highlight the importance of modeling survey-specific characteristics and cross-survey interactions, and provide guidance for building scalable classifiers for future time-domain astronomy.

Citations

• Leveraging pre-trained vision Transformers for multi-band photometric light curve classification
• Transfer Learning for Transient Classification: From Simulations to Real Data and ZTF to LSST
• Multiband Embeddings of Light Curves
• AstroM³: A self-supervised multimodal model for astronomy
• Transformer-Based Astronomical Time Series Model with Uncertainty Estimation for Detecting Misclassified Instances
• Maven: A Multimodal Foundation Model for Supernova Science
• Physical Properties of Type II Supernovae Inferred from ZTF and ATLAS Photometric Data
• Preliminary Report on Mantis Shrimp: a Multi-Survey Computer Vision Photometric Redshift Model
• A New Forced Photometry Service for the Zwicky Transient Facility
• Stellar Variability in Ground-Based Photometric Surveys: An Overview
• Astroconformer: Inferring Surface Gravity of Stars from Stellar Light Curves with Transformer
• Gaia Data Release 3
• Swin Transformer V2: Scaling Up Capacity and Resolution
• Transformer Tracking
• MaX-DeepLab: End-to-End Panoptic Segmentation with Mask Transformers
• End-to-End Object Detection with Transformers
• The Zwicky Transient Facility: System Overview, Performance, and First Results
• An Early Warning System for Asteroid Impact
• Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
• Swin Transformer: Hierarchical Vision Transformer using Shifted Windows

Related

• Leveraging pre-trained vision Transformers for multi-band photometric light curve classification
• Paying Attention to Astronomical Transients: Introducing the Time-series Transformer for Photometric Classification
• Multivariate Time-series Transformer Embeddings for Light Curves
• Photometric light curves classification with machine learning
• Real-time Light Curve Classification Framework for the Wide Field Survey Telescope Using Modified Semi-supervised Variational Auto-Encoder