Learning What's Real: Disentangling Signal and Measurement Artifacts in Multi-Sensor Data, with Applications to Astrophysics
Massachusetts Institute of Technology(麻省理工学院);Flatiron Institute, Simons Foundation(Flatiron研究所,Simons基金会);Institute for Advanced Studies(高级研究 institute);Harvard University(哈佛大学);New York University(纽约大学);Instituto de Astrofísica de Canarias(加那利大天文台)
Pablo Mercader-Perez, Carolina Cuesta-Lazaro, Daniel Muthukrishna, Jeroen Audenaert, V. Ashley Villar, David W. Hogg, Marc Huertas-Company, William T. Freeman
Data collected from the physical world is always a combination of multiple sources: an underlying signal from the physical process of interest and a signal from measurement-dependent artifacts from the sensor or instrument. This secondary signal acts as a confounding factor, limiting our ability to extract information about the physics underlying the phenomena we observe. Furthermore, it complicates the combination of observations in heterogeneous or multi-instrument settings. We propose a deep learning framework that leverages overlapping observations, a dual-encoder architecture, and a counterfactual generation objective to disentangle these factors of variation. The resulting representations explicitly separate intrinsic signals from sensor-specific distortions and noise, and can be used for counterfactual view generation, parameter inference unconfounded by measurement distortions, and instrument-independent similarity search. We demonstrate the effectiveness of our approach on astrophysical galaxy images from the DESI Legacy Imaging Survey (Legacy) and the Hyper Suprime-Cam (HSC) Survey as a representative multi-instrument setting. This framework provides a general recipe for scientific and multi-modal self-supervised pretraining: construct training pairs from overlapping observations of the same physical system, treat sensor- or modality-specific effects as augmentations, and learn invariant representations through counterfactual generation.
M-EPDet: Real-Time Real-Bogus Classification and Transient Candidate Judgement for the EP-WXT Pipeline via Multi-Modal Data
National Astronomical Observatories, Chinese Academy of Sciences(中国科学院国家天文台);University of Chinese Academy of Sciences(中国科学院大学);National Astronomical Data Center(国家天文数据中心)
Identification and mitigation of narrow spectral artifacts that degrade searches for persistent gravitational waves in the first two observing runs of Advanced LIGO
Tracking turbulence for wind-driven halo removal using MagAO-X pyramid wavefront sensor images
Steward Observatory, University of Arizona(亚利桑那大学斯图尔德天文台);James C. Wyant College of Optical Sciences, University of Arizona(亚利桑那大学詹姆斯·C·怀恩特光学科学学院);Center for Computational Astrophysics, Flatiron Institute(平顿研究所计算天体物理中心);Earth and Planets Laboratory, Carnegie Institution for Science(卡内基科学机构地球与行星实验室);Leiden Observatory, Leiden University(莱顿大学莱顿天文台);Subaru Telescope, National Observatory of Japan(日本国立天文台昴星团望远镜);Department of Astronomy, University of Michigan(密歇根大学天文学系);Northrop Grumman Corporation(诺斯罗普·格鲁曼公司);Draper Laboratory(德拉珀实验室);Starfire Optical Range, Kirtland Air Force Base(基尔兰德空军基地星光光学靶场)
WindsoCC: reconstructing the wind-driven halo in MagAO-X images using wavefront sensor telemetry
Steward Observatory, University of Arizona(亚利桑那大学斯图尔德天文台);James C. Wyant College of Optical Sciences, University of Arizona(亚利桑那大学詹姆斯·C·怀恩特光学科学学院);Center for Computational Astrophysics, Flatiron Institute(平顿研究所计算天体物理学中心);Earth and Planets Laboratory, Carnegie Institution for Science(卡内基科学机构地球与行星实验室);Leiden Observatory, Leiden University(莱顿大学莱顿天文台);Subaru Telescope, National Observatory of Japan(日本国立天文台昴星团望远镜);Department of Astronomy, University of Michigan(密歇根大学天文学系);Northrop Grumman Corporation(诺斯罗普·格鲁曼公司);Draper Laboratory(德雷珀实验室);Starfire Optical Range, Kirtland Air Force Base(基尔兰德空军基地星光光学靶场)
DeepExtractor: Time-domain reconstruction of signals and glitches in gravitational wave data with deep learning
Faculty of Science, Open Universiteit(科学学院,开放大学);Institute for Gravitational and Subatomic Physics (GRASP), Utrecht University(引力与亚原子物理研究所(GRASP),乌得勒支大学);Nikhef(Nikhef研究所);Department of Physics, University of Massachusetts(物理系,马萨诸塞大学)