5 papers
ASTROCO: Self-Supervised Conformer-Style Transformers for Light-Curve Embeddings
Antony Tan, Pavlos Protopapas, Martina Cádiz-Leyton +3
We present AstroCo, a Conformer-style encoder for irregular stellar light curves. By combining attention with depthwise convolutions and gating, AstroCo captures both global depend…
Astro-MoE: Mixture of Experts for Multiband Astronomical Time Series
Martina Cádiz-Leyton, Guillermo Cabrera-Vives, Pavlos Protopapas +2
Multiband astronomical time series exhibit heterogeneous variability patterns, sampling cadences, and signal characteristics across bands. Standard transformers apply shared parame…
Leveraging pre-trained vision Transformers for multi-band photometric light curve classification
Daniel Moreno-Cartagena, Pavlos Protopapas, Guillermo Cabrera-Vives +3
This study investigates the potential of a pre-trained vision Transformer (VT) model, specifically the Swin Transformer V2 (SwinV2), to classify photometric light curves without th…
Multiband Embeddings of Light Curves
I. Becker, P. Protopapas, M. Catelan +1
In this work, we propose a novel ensemble of recurrent neural networks (RNNs) that considers the multiband and non-uniform cadence without having to compute complex features. Our p…
Uncertainty estimation for time series classification: Exploring predictive uncertainty in transformer-based models for variable stars
Martina Cádiz-Leyton, Guillermo Cabrera-Vives, Pavlos Protopapas +3
Classifying variable stars is key for understanding stellar evolution and galactic dynamics. With the demands of large astronomical surveys, machine learning models, especially att…