A Multimodal Dataset and Benchmark for Radio Galaxy and Infrared Host Detection
Computer Vision and Pattern Recognition2023-12-13v1Cosmology and Nongalactic AstrophysicsAstrophysics of GalaxiesInstrumentation and Methods for Astrophysics
We present a novel multimodal dataset developed by expert astronomers to automate the detection and localisation of multi-component extended radio galaxies and their corresponding infrared hosts. The dataset comprises 4,155 instances of galaxies in 2,800 images with both radio and infrared modalities. Each instance contains information on the extended radio galaxy class, its corresponding bounding box that encompasses all of its components, pixel-level segmentation mask, and the position of its corresponding infrared host galaxy. Our dataset is the first publicly accessible dataset that includes images from a highly sensitive radio telescope, infrared satellite, and instance-level annotations for their identification. We benchmark several object detection algorithms on the dataset and propose a novel multimodal approach to identify radio galaxies and the positions of infrared hosts simultaneously.
@article{arxiv.2312.06728,
title = {A Multimodal Dataset and Benchmark for Radio Galaxy and Infrared Host Detection},
author = {Nikhel Gupta and Zeeshan Hayder and Ray P. Norris and Minh Hyunh and Lars Petersson},
journal= {arXiv preprint arXiv:2312.06728},
year = {2023}
}
Comments
Accepted in NeurIPS 2023 conference ML4PS workshop (https://nips.cc/). The full version accepted in PASA, is available at https://doi.org/10.1017/pasa.2023.64