Electromagnetic (EM) body models designed to predict Radio-Frequency (RF)
propagation are time-consuming methods which prevent their adoption in strict
real-time computational imaging problems, such as human body localization and
sensing. Physics-informed Generative Neural Network (GNN) models have been
recently proposed to reproduce EM effects, namely to simulate or reconstruct
missing data or samples by incorporating relevant EM principles and
constraints. The paper discusses a Variational Auto-Encoder (VAE) model which
is trained to reproduce the effects of human motions on the EM field and
incorporate EM body diffraction principles. Proposed physics-informed
generative neural network models are verified against both classical
diffraction-based EM tools and full-wave EM body simulations.

该研究论文讨论了一种基于变分自编码器（VAE）的物理约束生成神经网络模型，用于模拟人体运动对电磁场的影响并融合电磁体绕射原理。通过与经典绕射基于电磁工具和全波电磁体模拟进行验证。

应用物理信息的生成神经网络用于室内人体感知的射频传播预测

Physics-informed generative neural networks for RF propagation  prediction with application to indoor body perception

Combined gravitational-wave (GW) and electromagnetic (EM) observations of
compact binary mergers should enable detailed studies of astrophysical
processes in the strong-field gravity regime. Networks of GW interferometers
have poor angular resolution on the sky and their EM signatures are predicted
to be faint. Therefore, a challenging goal will be to unambiguously pinpoint
the EM counterparts to GW mergers. We perform the first comprehensive
end-to-end simulation that focuses on: i) GW sky localization, distance
measures and volume errors with two compact binary populations and four
different GW networks, ii) subsequent detectability by a slew of
multiwavelength telescopes and, iii) final identification of the merger
counterpart amidst a sea of possible astrophysical false-positives. First, we
find that double neutron star (NS) binary mergers can be detected out to a
maximum distance of 400 Mpc (or 750 Mpc) by three (or five) detector GW
networks respectively. NS -- black-hole (BH) mergers can be detected a factor
of 1.5 further out. The sky localization uncertainties for NS-BH mergers are
50--170 sq. deg. (or 6--65 sq. deg.) for a three (or five detector) GW network
respectively. Second, we quantify relative fractions of optical counterparts
that are detectable by different size telescopes. Third, we present five case
studies to illustrate the diversity of challenges in secure identification of
the EM counterpart at low and high Galactic latitudes. For the first time, we
demonstrate how construction of low-latency GW volumes in conjunction with
local universe galaxy catalogs can help solve the problem of false positives.

本研究模拟 GW 和 EM 观测点源的情况，定量地研究了 EM 对应物探测的挑战性，并提出了构建低延迟 GW 体积和当地星系目录相结合，以解决错误阳性问题的方法。