Logo image
Back
DeepVoid: A Deep Learning Void Detector
Journal article   Open access   Peer reviewed

DeepVoid: A Deep Learning Void Detector

Sam Kumagai, Michael S. Vogeley, Miguel A. Aragon-Calvo, Kelly A. Douglass, Segev BenZvi and Mark Neyrinck
The Astrophysical journal, v 998(1), 85
10 Feb 2026
DOI: https://doi.org/10.3847/1538-4357/ae2c80
url
https://doi.org/10.3847/1538-4357/ae2c80View
Published, Version of Record (VoR) Open CC BY V4.0

Abstract

Astronomy & Astrophysics Science & Technology Physical Sciences
We present DeepVoid, an application of deep learning trained on a physical definition of cosmic voids to detect voids in density fields and galaxy distributions. By semantically segmenting the IllustrisTNG simulation volume using the tidal tensor, we train a deep convolutional neural network to classify local structure using a U-Net architecture for training and prediction. The model achieves a void F1 score of 0.96 and a Matthews correlation coefficient over all structural classes of 0.81 for dark matter particles in IllustrisTNG with interparticle spacing of lambda = 0.33 h-1 Mpc. We then apply the machine learning technique of curriculum learning to enable the model to classify structure in data with significantly larger intertracer separation. At the highest tracer separation tested, lambda = 10 h-1 Mpc, the model achieves a void F1 score of 0.89 and a Matthews correlation coefficient of 0.6 on IllustrisTNG subhalos.

Metrics

1 Record Views

Details

Logo image