Transforming a rare event search into a not-so-rare event search in real-time with deep learning-based object detection

Deep learning-based object detection algorithms enable the simultaneous classification and localization of any number of objects in image data. Many of these algorithms are capable of operating in real-time on high resolution images, attributing to their widespread usage across many fields. We present an end-to-end object detection pipeline designed for rare event searches for the Migdal effect, at real-time speeds, using high-resolution image data from the scientific CMOS camera readout of the MIGDAL experiment. The Migdal effect in nuclear scattering, critical for sub-GeV dark matter searches, has yet to be experimentally confirmed, making its detection a primary goal of the MIGDAL experiment. The Migdal effect forms a composite rare event signal topology consisting of an electronic and nuclear recoil sharing the same vertex. Crucially, both recoil species are commonly observed in isolation in the MIGDAL experiment, enabling us to train YOLOv8, a state-of-the-art object detection algorithm, on real data. Topologies indicative of the Migdal effect can then be identified in science data via pairs of neighboring or overlapping electron and nuclear recoils. Applying selections to real data that retain 99.7% signal acceptance in simulations, we demonstrate our pipeline to reduce a sample of 20 million recorded images to fewer than 1000 frames, thereby transforming a rare search into a much more manageable search. More broadly, we discuss the applicability of using object detection to enable data-driven machine learning training for other rare event search applications such as neutrinoless double beta decay searches and experiments imaging exotic nuclear decays.

Authors

Institutions

Publication Details

Journal
Physical review. D/Physical review. D.
Published
2025-04-09
DOI
https://doi.org/10.1103/physrevd.111.072004
Citations
3
Primary Topic
Radiation Detection and Scintillator Technologies
Type
article
Field-Weighted Citation Impact
5.65

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Transforming a rare event search into a not-so-rare event search in real-time with deep learning-based object detection

M. Nakhostin, M. Lisowska, F. Brunbauer, K. Nikolopoulos et al.
3 citations
Physical review. D/Physical review. D.
Radiation Detection and Scintillator Technologies
5.65
article

Transforming a rare event search into a not-so-rare event search in real-time with deep learning-based object detection

M. Nakhostin, M. Lisowska, F. Brunbauer, K. Nikolopoulos, V. N. Solovov, P. Knights, Christopher Frost, R. Veenhof, L. Ropelewski, S. Lilley, I. Katsioulas, V. A. Kudryavtsev, A. F. Mills, Maria Kastriotou, Christopher McCabe, A. Cottle, I. Lopes, S. N. Balashov, D. Loomba, E. Tilly, P. Luna Dapica, F. Neves, A. C. Kaboth, H. M. Araújo, T. J. Neep, J. Schueler, D. Hunt, A. Lindote, R. Turnley, T. Marley, T. J. Sumner, L. Millins, J. Borg, E. Oliveri, C. Cazzaniga, F. García, H. Kraus, R. Nandakumar, C. Brew, E. López Asamar, P. A. Majewski, J. Tarrant, A. Khazov
article en
3 citations

Abstract

Deep learning-based object detection algorithms enable the simultaneous classification and localization of any number of objects in image data. Many of these algorithms are capable of operating in real-time on high resolution images, attributing to their widespread usage across many fields. We present an end-to-end object detection pipeline designed for rare event searches for the Migdal effect, at real-time speeds, using high-resolution image data from the scientific CMOS camera readout of the MIGDAL experiment. The Migdal effect in nuclear scattering, critical for sub-GeV dark matter searches, has yet to be experimentally confirmed, making its detection a primary goal of the MIGDAL experiment. The Migdal effect forms a composite rare event signal topology consisting of an electronic and nuclear recoil sharing the same vertex. Crucially, both recoil species are commonly observed in isolation in the MIGDAL experiment, enabling us to train YOLOv8, a state-of-the-art object detection algorithm, on real data. Topologies indicative of the Migdal effect can then be identified in science data via pairs of neighboring or overlapping electron and nuclear recoils. Applying selections to real data that retain 99.7% signal acceptance in simulations, we demonstrate our pipeline to reduce a sample of 20 million recorded images to fewer than 1000 frames, thereby transforming a rare search into a much more manageable search. More broadly, we discuss the applicability of using object detection to enable data-driven machine learning training for other rare event search applications such as neutrinoless double beta decay searches and experiments imaging exotic nuclear decays.

Physical review. D/Physical review. D.Vol. 111(7)
Rutherford Appleton Laboratory (GB), University of Helsinki (FI), Universität Hamburg (DE), King's College London (GB), Royal Holloway University of London (GB), Luleå University of Technology (SE), Université Paris-Saclay (FR), University of Oxford (GB), ISIS Neutron and Muon Source, Imperial College London (GB), Universidad Autónoma de Madrid (ES), European Organization for Nuclear Research (CH), University of Coimbra (PT), University of Birmingham (GB), University of Sheffield (GB)
National Science Foundation, U.S. Department of Energy, CERN, UK Research and Innovation, European Commission, Deutsche Forschungsgemeinschaft, Universidad Autónoma de Madrid, Office of Science, Horizon 2020 Framework Programme, Science and Technology Facilities Council, Fundação para a Ciência e a Tecnologia, High Energy Physics, H2020 Marie Skłodowska-Curie Actions
Openalex Percentile: Top 4%
Radiation Detection and Scintillator Technologies
5.65
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.