arXiv subjects

Suhyeon Kim

Publications and source records attributed to Suhyeon Kim.

6 recordsLinked to original sources

arXiv2026

NeuVolEx: Implicit Neural Features for Volume Exploration

Direct volume rendering (DVR) aims to help users identify and examine regions of interest (ROIs) within volumetric data, and feature representations that support effective ROI classification and clustering play a fundamental role in volume exploration. Existing approaches typically rely on either explicit local feature representations or implicit convolutional feature representations learned from raw volumes. However, explicit local feature representations are limited in capturing broader geometric patterns and spatial correlations, while implicit convolutional feature representations do not necessarily ensure robust performance in practice, where user supervision is typically limited. Meanwhile, implicit neural representations (INRs) have recently shown strong promise in DVR for volume compression, owing to their ability to compactly parameterize continuous volumetric fields. In this work, we propose NeuVolEx, a neural volume exploration approach that extends the role of INRs beyond volume compression. Unlike prior compression methods that focus on INR outputs, NeuVolEx leverages feature representations learned during INR training as a robust basis for volume exploration. To better adapt these feature representations to exploration tasks, we augment a base INR with a structural encoder and a multi-task learning scheme that improve spatial coherence for ROI characterization. We validate NeuVolEx on two fundamental volume exploration tasks: image-based transfer function (TF) design and viewpoint recommendation. NeuVolEx enables accurate ROI classification under sparse user supervision for image-based TF design and supports unsupervised clustering to identify compact complementary viewpoints that reveal different ROI clusters. Experiments on diverse volume datasets with varying modalities and ROI complexities demonstrate NeuVolEx improves both effectiveness and usability over prior methods

cs.GR↗

arXiv2025

A comparative study of physics capabilities of a liquid argon and a water based liquid scintillator at DUNE

We present a comprehensive comparison of the physics sensitivities of a Liquid Argon Time Projection Chamber (LArTPC) and a Water-based Liquid Scintillator (WbLS) detector, considering their potential deployment as the fourth far detector module in the DUNE facility. Using GLoBES-based simulations, we evaluate their performance in measuring standard neutrino oscillation parameters ($θ_{23}, δ_{13}$ and $Δm^{2}_{31}$), both in standard 3-neutrino case, as well as in presence of new physics scenarios involving light sterile neutrinos and neutral-current non-standard interactions (NC NSI). Our findings show that THEIA (a WbLS-based detector) significantly outperforms LArTPC in resolving the CP phase $δ_{13}$,- especially near maximal CP violation, and in lifting the octant degeneracy of $θ_{23}$ due to its superior energy resolution and ability to clearly identify the second oscillation maximum. Furthermore, THEIA offers competitive reconstruction precision even with relatively moderate energy resolutions ($7-10\%/\sqrt{E}$) and demonstrates enhanced robustness under new physics scenarios. These results support the physics-driven case for a hybrid DUNE configuration utilizing both LArTPC and WbLS technologies for optimized sensitivity across the full spectrum of neutrino oscillation and physics beyond the standard model.

hep-ph↗

arXiv2025

Review learning: Real world validation of privacy preserving continual learning across medical institutions

When a deep learning model is trained sequentially on different datasets, it often forgets the knowledge learned from previous data, a problem known as catastrophic forgetting. This damages the model's performance on diverse datasets, which is critical in privacy-preserving deep learning (PPDL) applications based on transfer learning (TL). To overcome this, we introduce "review learning" (RevL), a low cost continual learning algorithm for diagnosis prediction using electronic health records (EHR) within a PPDL framework. RevL generates data samples from the model which are used to review knowledge from previous datasets. Six simulated institutional experiments and one real-world experiment involving three medical institutions were conducted to validate RevL, using three binary classification EHR data. In the real-world experiment with data from 106,508 patients, the mean global area under the receiver operating curve was 0.710 for RevL and 0.655 for TL. These results demonstrate RevL's ability to retain previously learned knowledge and its effectiveness in real-world PPDL scenarios. Our work establishes a realistic pipeline for PPDL research based on model transfers across institutions and highlights the practicality of continual learning in real-world medical settings using private EHR data.

cs.AI↗

arXiv2024

Probing Large Extra Dimension at DUNE using beam tunes

The Deep Underground Neutrino Experiment (DUNE) is a leading experiment in neutrino physics which is presently under construction. DUNE aims to measure the yet unknown parameters in the three flavor oscillation case which includes discovery of leptonic CP violation, determination of the neutrino mass hierarchy and measuring the octant of $θ_{23}$. Additionally, the ancillary goals of DUNE include probing the subdominant effects induced by possible physics beyond the Standard Model (BSM). One such new physics scenario is the possible presence of Large Extra Dimension (LED) which can naturally give rise to tiny neutrino masses. LED impacts neutrino oscillation through two new parameters, - namely the lightest Dirac mass $m_{0}$ and the radius of the extra dimension $R_{\text{ED}}$ ($< 2$ $μ$m). At the DUNE baseline of 1300 km, the probability seems to be modified more at the higher energy ($\gtrsim 4-5$ GeV) in presence of LED. In this work, we attempt to constrain the parameter space of $m_{0}$ and $R_{\text{ED}}$ by performing a statistical analysis of neutrino data simulated at DUNE far detector (FD). We illustrate how a combination of the standard low energy (LE) neutrino beam and a medium energy (ME) neutrino beam can take advantage of the relatively large impact of LED at higher energy and improve the constraints. In the analysis we also show the role of the individual oscillation channels ($ν_μ \to ν_{e}, ν_μ \to ν_μ, ν_μ \to ν_τ$), as well as the two neutrino mass hierarchies.

hep-ph↗

arXiv2022

Photoluminescence Path Bifurcations by Spin Flip in Two-Dimensional CrPS4

Ultrathin layered crystals of coordinated chromium(III) are promising not only as two-dimensional (2D) magnets but also as 2D near-infrared (NIR) emitters owing to long-range spin correlation and efficient transition between high and low-spin excited states of Cr3+ ions. In this study, we report on dual-band NIR photoluminescence (PL) of CrPS4 and show that its excitonic emission bifurcates into fluorescence and phosphorescence depending on thickness, temperature and defect density. In addition to the spectral branching, the biexponential decay of PL transients, also affected by the three factors, could be well described within a three-level kinetic model for Cr(III). In essence, the PL bifurcations are governed by activated reverse intersystem crossing from the low to high-spin states, and the transition barrier becomes lower for thinner 2D samples because of surface-localized defects. Our findings can be generalized to 2D solids of coordinated metals and will be valuable in realizing novel magneto-optic functions and devices.

cond-mat.mtrl-sci↗

arXiv2020

Crossover between Photochemical and Photothermal Oxidations of Atomically Thin Magnetic Semiconductor CrPS4

Many two-dimensional (2D) semiconductors represented by transition metal dichalcogenides have tunable optical bandgaps in the visible or near IR-range standing as a promising candidate for optoelectronic devices. Despite this potential, however, their photoreactions are not well understood or controversial in the mechanistic details. In this work, we report a unique thickness-dependent photoreaction sensitivity and a switchover between two competing reaction mechanisms in atomically thin chromium thiophosphate (CrPS4), a two-dimensional antiferromagnetic semiconductor. CrPS4 showed a threshold power density 2 orders of magnitude smaller than that for MoS2 obeying a photothermal reaction route. In addition, reaction cross section quantified with Raman spectroscopy revealed distinctive power dependences in the low and high power regimes. On the basis of optical in situ thermometric measurements and control experiments against O2, water, and photon energy, we proposed a photochemical oxidation mechanism involving singlet O2 in the low power regime with a photothermal route for the other. We also demonstrated a highly effective encapsulation with Al2O3 as a protection against the destructive photoinduced and ambient oxidations.

cond-mat.mtrl-sci↗