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

TL;DR

This paper demonstrates that using a pre-trained vision Transformer to jointly process multi-survey light curves improves classification accuracy in time-domain astronomy, emphasizing the importance of survey-specific modeling.

Contribution

It introduces a multi-survey classification approach with a pre-trained vision Transformer, highlighting the benefits of joint processing of diverse survey data.

Findings

Joint survey processing outperforms single-survey models.

Modeling survey-specific features enhances classification.

Guidelines for scalable multi-survey classifiers are provided.

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.