- Home
- Volume 93 - Année 2024
- No 3 - 41st Liège International Astrophysical Coll...
- Adaptation of a Convolutional Neural Network–based Pipeline to Detect Short Gravitational Wave Bursts
View(s): 12 (4 ULiège)
Download(s): 0 (0 ULiège)
Adaptation of a Convolutional Neural Network–based Pipeline to Detect Short Gravitational Wave Bursts
Attached document(s)
original pdf fileAbstract
We present a machine learning based pipeline to analyze unmodeled gravitational wave (GW) transients of less than 10 s. The convolutional neural network (CNN) is based on a U-NET architecture and takes as input data from GW interferometers represented as time-frequency maps, returning a spectrogram without the background noise. The CNN has been trained on simulated data, using a generated Gaussian background noise and injecting GW signals from core-collapse supernovae (CCSNe) simulations. The pipeline is able to successfully denoise spectrograms and recognize as signals also CCSNe waveforms for which it has not been trained on.
This work is distributed under the Creative Commons CC BY 4.0 Licence.
Paper presented at the 41st Liège International Astrophysical Colloquium on “The eventful life of massive star multiples,” University of Liège (Belgium), 15–19 July 2024.
Bibliographie
[1] Jebur, R. S., Zabil, M. H. B. M., Hammood, D. A., and Cheng, L. K. (2024) A comprehensive review of image denoising in deep learning. Multimedia Tools and Applications, 83(20), 58 181–58 199. https://doi.org/10.1007/s11042-023-17468-2.
[2] Boudart, V. and Fays, M. (2022) Machine learning algorithm for minute-long burst searches. PhRvD, 105(8), 083007. https://doi.org/10.1103/PhysRevD.105.083007.
[3] Boudart, V. (2023) Detection of minute-long Gravitational Wave transients using Deep Learning methods. Ph.D. thesis, Université de Liège [Sciences], Liège (BE). https://hdl.handle.net/2268/308320.
[4] Nitz, A., Harry, I., Brown, D., Biwer, C. M., Willis, J., Canton, T. D., Capano, C., Dent, T., Pekowsky, L., Davies, G. S. C., De, S., Cabero, M., Wu, S., Williamson, A. R., Machenschalk, B., Macleod, D., Pannarale, F., Kumar, P., Reyes, S., dfinstad, Kumar, S., Tápai, M., Singer, L., Kumar, P., veronica-villa, maxtrevor, Gadre, B. U. V., Khan, S., Fairhurst, S., and Tolley, A. (2024). gwastro/pycbc: v2.3.3 release of pycbc. Zenodo [Software]. https://doi.org/10.5281/zenodo.10473621.
[5] Xing, F., Xie, Y., Shi, X., Chen, P., Zhang, Z., and Yang, L. (2019) Towards pixel-to-pixel deep nucleus detection in microscopy images. BMC Bioinformatics, 20. https://doi.org/10.1186/s12859-019-3037-5.
[6] Powell, J. and Müller, B. (2019) Gravitational wave emission from 3D explosion models of core-collapse supernovae with low and normal explosion energies. MNRAS, 487(1), 1178–1190. https://doi.org/10.1093/mnras/stz1304.
[7] Radice, D., Morozova, V., Burrows, A., Vartanyan, D., and Nagakura, H. (2019) Characterizing the gravitational wave signal from core-collapse supernovae. ApJL, 876(1), L9. https://doi.org/10.3847/2041-8213/ab191a.
[8] O’Connor, E. P. and Couch, S. M. (2018) Exploring fundamentally three-dimensional phenomena in high-fidelity simulations of core-collapse supernovae. ApJ, 865(2), 81. https://doi.org/10.3847/1538-4357/aadcf7.
[9] Powell, J. and Müller, B. (2020) Three-dimensional core-collapse supernova simulations of massive and rotating progenitors. MNRAS, 494(4), 4665–4675. https://doi.org/10.1093/mnras/staa1048.
[10] Boudart, V. (2023) Convolutional neural network to distinguish glitches from minute-long gravitational wave transients. PhRvD, 107(2), 024007. https://doi.org/10.1103/PhysRevD.107.024007.
To cite this article
About: Matteo Pracchia
email : mpracchia@uliege.be