Searcharxiv⌕ Search

arXiv subjects

Hyojin Kim

Publications and source records attributed to Hyojin Kim.

At least 19 recordsLinked to original sources
arXiv2026

Sharing a Fabric with Collective Communication: Two Storage Penalties in Deep Learning Training

Distributed DL training on HPC systems often shares one network fabric between NCCL/RCCL collective communication and parallel-filesystem I/O. Using a real GNN training workload on a Slingshot-11 system, we show that this sharing imposes two distinct costs. The primary cost is heavy-tailed DataLoader stalls: the typical DataLoader wait is just 15 ms at steady state, yet spikes to multiple seconds in 28% of Lustre iterations and 12% of VAST iterations. The secondary cost is traffic-class contention on collective communication: Lustre I/O stalls the all-reduce by up to 145$\times$ in an isolated benchmark. The two costs arise from different mechanisms. I/O stall latency affects any storage path that traverses the shared fabric, whereas all-reduce network contention occurs only when storage and collective communication share the same traffic class. Their common root cause is that storage I/O traverses the shared fabric. This work shows that node-local NVMe staging via DYAD (Our code is publicly available at https://github.com/flux-framework/dyad) eliminates both effects by keeping storage I/O off that path. Across a full training epoch, DYAD achieves a 7.4 times speedup over direct Lustre reads and a 1.06 times speedup over VAST. By the second epoch, once the local cache is fully warmed, DataLoader stalls are eliminated entirely, allowing DYAD to reach a 1.31 times speedup over VAST.

cs.DC↗
arXiv2026

Does Persona Make LLMs K-pop Fans? A Pilot Study of LLM-Based Online Concert Audience Agents

A concert is a collective experience, but recorded performance videos are typically watched alone, stripping away the shared audience presence that makes concerts feel eventful. We investigate whether persona-based LLM audience agents can recreate aspects of this collective experience by generating real-time fan chat alongside a K-pop performance video. We present a multi-agent system in which ten LLM agents react through live-chat messages, comparing a persona-conditioned audience (each agent assigned a distinct fan identity, bias, and chat style) with a no-persona baseline. In a within-subjects pilot with K-pop fans (N=11), persona conditioning substantially improved model-level chat quality and perceived naturalness, but did not translate into differences in social connectedness, engagement, or affective response. Interviews suggest that online K-pop concert chat may operate as collective monologue rather than interpersonal dialogue, and that meaningful participation depends on shared identification with the specific artist and fandom. Persona conditioning can make LLM audiences appear more natural, but culturally meaningful collective experience may require deeper alignment between persona, crowd behavior, fandom identity, and user expectations.

cs.HC↗
arXiv2026

Unpaired Image-to-Image Translation via a Self-Supervised Semantic Bridge

Adversarial diffusion and diffusion-inversion methods have advanced unpaired image-to-image translation, but each faces key limitations. Adversarial approaches require target-domain adversarial loss during training, which can limit generalization to unseen data, while diffusion-inversion methods often produce low-fidelity translations due to imperfect inversion into noise-latent representations. In this work, we propose the Self-Supervised Semantic Bridge (SSB), a versatile framework that integrates external semantic priors into diffusion bridge models to enable spatially faithful translation without cross-domain supervision. Our key idea is to leverage self-supervised visual encoders to learn representations that are invariant to appearance changes but capture geometric structure, forming a shared latent space that conditions the diffusion bridges. Extensive experiments show that SSB outperforms strong prior methods for challenging medical image synthesis in both in-domain and out-of-domain settings, and extends easily to high-quality text-guided editing.

cs.CV↗
arXiv2024

Bimodal Visualization of Industrial X-Ray and Neutron Computed Tomography Data

Advanced manufacturing creates increasingly complex objects with material compositions that are often difficult to characterize by a single modality. Our collaborating domain scientists are going beyond traditional methods by employing both X-ray and neutron computed tomography to obtain complementary representations expected to better resolve material boundaries. However, the use of two modalities creates its own challenges for visualization, requiring either complex adjustments of bimodal transfer functions or the need for multiple views. Together with experts in nondestructive evaluation, we designed a novel interactive bimodal visualization approach to create a combined view of the co-registered X-ray and neutron acquisitions of industrial objects. Using an automatic topological segmentation of the bivariate histogram of X-ray and neutron values as a starting point, the system provides a simple yet effective interface to easily create, explore, and adjust a bimodal visualization. We propose a widget with simple brushing interactions that enables the user to quickly correct the segmented histogram results. Our semiautomated system enables domain experts to intuitively explore large bimodal datasets without the need for either advanced segmentation algorithms or knowledge of visualization techniques. We demonstrate our approach using synthetic examp

cs.GR↗
arXiv2024

Generalizing AI-driven Assessment of Immunohistochemistry across Immunostains and Cancer Types: A Universal Immunohistochemistry Analyzer

Despite advancements in methodologies, immunohistochemistry (IHC) remains the most utilized ancillary test for histopathologic and companion diagnostics in targeted therapies. However, objective IHC assessment poses challenges. Artificial intelligence (AI) has emerged as a potential solution, yet its development requires extensive training for each cancer and IHC type, limiting versatility. We developed a Universal IHC (UIHC) analyzer, an AI model for interpreting IHC images regardless of tumor or IHC types, using training datasets from various cancers stained for PD-L1 and/or HER2. This multi-cohort trained model outperforms conventional single-cohort models in interpreting unseen IHCs (Kappa score 0.578 vs. up to 0.509) and consistently shows superior performance across different positive staining cutoff values. Qualitative analysis reveals that UIHC effectively clusters patches based on expression levels. The UIHC model also quantitatively assesses c-MET expression with MET mutations, representing a significant advancement in AI application in the era of personalized medicine and accumulating novel biomarkers.

cs.CV↗
arXiv2024

Advancing Cross-Domain Generalizability in Face Anti-Spoofing: Insights, Design, and Metrics

This paper presents a novel perspective for enhancing anti-spoofing performance in zero-shot data domain generalization. Unlike traditional image classification tasks, face anti-spoofing datasets display unique generalization characteristics, necessitating novel zero-shot data domain generalization. One step forward to the previous frame-wise spoofing prediction, we introduce a nuanced metric calculation that aggregates frame-level probabilities for a video-wise prediction, to tackle the gap between the reported frame-wise accuracy and instability in real-world use-case. This approach enables the quantification of bias and variance in model predictions, offering a more refined analysis of model generalization. Our investigation reveals that simply scaling up the backbone of models does not inherently improve the mentioned instability, leading us to propose an ensembled backbone method from a Bayesian perspective. The probabilistically ensembled backbone both improves model robustness measured from the proposed metric and spoofing accuracy, and also leverages the advantages of measuring uncertainty, allowing for enhanced sampling during training that contributes to model generalization across new datasets. We evaluate the proposed method from the benchmark OMIC dataset and also the public CelebA-Spoof and SiW-Mv2. Our final model outperforms existing state-of-the-art methods across the datasets, showcasing advancements in Bias, Variance, HTER, and AUC metrics.

cs.CV↗
arXiv2024

Distributed Stochastic Optimization of a Neural Representation Network for Time-Space Tomography Reconstruction

4D time-space reconstruction of dynamic events or deforming objects using X-ray computed tomography (CT) is an important inverse problem in non-destructive evaluation. Conventional back-projection based reconstruction methods assume that the object remains static for the duration of several tens or hundreds of X-ray projection measurement images (reconstruction of consecutive limited-angle CT scans). However, this is an unrealistic assumption for many in-situ experiments that causes spurious artifacts and inaccurate morphological reconstructions of the object. To solve this problem, we propose to perform a 4D time-space reconstruction using a distributed implicit neural representation (DINR) network that is trained using a novel distributed stochastic training algorithm. Our DINR network learns to reconstruct the object at its output by iterative optimization of its network parameters such that the measured projection images best match the output of the CT forward measurement model. We use a forward measurement model that is a function of the DINR outputs at a sparsely sampled set of continuous valued 4D object coordinates. Unlike previous neural representation architectures that forward and back propagate through dense voxel grids that sample the object's entire time-space coordinates, we only propagate through the DINR at a small subset of object coordinates in each iteration resulting in an order-of-magnitude reduction in memory and compute for training. DINR leverages distributed computation across several compute nodes and GPUs to produce high-fidelity 4D time-space reconstructions. We use both simulated parallel-beam and experimental cone-beam X-ray CT datasets to demonstrate the superior performance of our approach.

eess.IV↗
arXiv2023

Segment-Based Wall Treatment Model for Heat Transfer Rate in Smoothed Particle Hydrodynamics

In this study, a smoothed particle hydrodynamics (SPH) model that applies a segment-based boundary treatment is used to simulate natural convection. In a natural convection simulated using an SPH model, the wall boundary treatment is a major issue because accurate heat transfer from boundaries should be calculated. The boundary particle method, which models the boundary by placing multiple layers of particles on and behind the wall boundary, is the most widely used boundary treatment method. Although this method can impose accurate boundary conditions, boundary modeling for complex shapes is challenging and requires excessive computational costs depending on the boundary shape. In this study, we utilize a segment-based boundary treatment method to model the wall boundary and apply this method to the energy conservation equation for the wall heat transfer model. The proposed method solves the problems arising from the use of boundary particles and simultaneously provides accurate heat transfer calculation results for the wall. In various numerical examples, the proposed method is verified through a comparison with available experimental results, SPH results using the boundary particle method, and finite volume method (FVM) results.

physics.flu-dyn↗
arXiv2023

Differentiable Forward Projector for X-ray Computed Tomography

Data-driven deep learning has been successfully applied to various computed tomographic reconstruction problems. The deep inference models may outperform existing analytical and iterative algorithms, especially in ill-posed CT reconstruction. However, those methods often predict images that do not agree with the measured projection data. This paper presents an accurate differentiable forward and back projection software library to ensure the consistency between the predicted images and the original measurements. The software library efficiently supports various projection geometry types while minimizing the GPU memory footprint requirement, which facilitates seamless integration with existing deep learning training and inference pipelines. The proposed software is available as open source: https://github.com/LLNL/LEAP.

cs.LG↗
arXiv2022

DOLCE: A Model-Based Probabilistic Diffusion Framework for Limited-Angle CT Reconstruction

Limited-Angle Computed Tomography (LACT) is a non-destructive evaluation technique used in a variety of applications ranging from security to medicine. The limited angle coverage in LACT is often a dominant source of severe artifacts in the reconstructed images, making it a challenging inverse problem. We present DOLCE, a new deep model-based framework for LACT that uses a conditional diffusion model as an image prior. Diffusion models are a recent class of deep generative models that are relatively easy to train due to their implementation as image denoisers. DOLCE can form high-quality images from severely under-sampled data by integrating data-consistency updates with the sampling updates of a diffusion model, which is conditioned on the transformed limited-angle data. We show through extensive experimentation on several challenging real LACT datasets that, the same pre-trained DOLCE model achieves the SOTA performance on drastically different types of images. Additionally, we show that, unlike standard LACT reconstruction methods, DOLCE naturally enables the quantification of the reconstruction uncertainty by generating multiple samples consistent with the measured data.

eess.IV↗
arXiv2022

Deep reinforcement learning for large-eddy simulation modeling in wall-bounded turbulence

The development of a reliable subgrid-scale (SGS) model for large-eddy simulation (LES) is of great importance for many scientific and engineering applications. Recently, deep learning approaches have been tested for this purpose using high-fidelity data such as direct numerical simulation (DNS) in a supervised learning process. However, such data are generally not available in practice. Deep reinforcement learning (DRL) using only limited target statistics can be an alternative algorithm in which the training and testing of the model are conducted in the same LES environment. The DRL of turbulence modeling remains challenging owing to its chaotic nature, high dimensionality of the action space, and large computational cost. In the present study, we propose a physics-constrained DRL framework that can develop a deep neural network (DNN)-based SGS model for the LES of turbulent channel flow. The DRL models that produce the SGS stress were trained based on the local gradient of the filtered velocities. The developed SGS model automatically satisfies the reflectional invariance and wall boundary conditions without an extra training process so that DRL can quickly find the optimal policy. Furthermore, direct accumulation of reward, spatially and temporally correlated exploration, and the pre-training process are applied for the efficient and effective learning. In various environments, our DRL could discover SGS models that produce the viscous and Reynolds stress statistics perfectly consistent with the filtered DNS. By comparing various statistics obtained by the trained models and conventional SGS models, we present a possible interpretation of better performance of the DRL model.

physics.flu-dyn↗
arXiv2021

High-Throughput Virtual Screening of Small Molecule Inhibitors for SARS-CoV-2 Protein Targets with Deep Fusion Models

Structure-based Deep Fusion models were recently shown to outperform several physics- and machine learning-based protein-ligand binding affinity prediction methods. As part of a multi-institutional COVID-19 pandemic response, over 500 million small molecules were computationally screened against four protein structures from the novel coronavirus (SARS-CoV-2), which causes COVID-19. Three enhancements to Deep Fusion were made in order to evaluate more than 5 billion docked poses on SARS-CoV-2 protein targets. First, the Deep Fusion concept was refined by formulating the architecture as one, coherently backpropagated model (Coherent Fusion) to improve binding-affinity prediction accuracy. Secondly, the model was trained using a distributed, genetic hyper-parameter optimization. Finally, a scalable, high-throughput screening capability was developed to maximize the number of ligands evaluated and expedite the path to experimental evaluation. In this work, we present both the methods developed for machine learning-based high-throughput screening and results from using our computational pipeline to find SARS-CoV-2 inhibitors.

cs.LG↗
arXiv2021

Dynamic CT Reconstruction from Limited Views with Implicit Neural Representations and Parametric Motion Fields

Reconstructing dynamic, time-varying scenes with computed tomography (4D-CT) is a challenging and ill-posed problem common to industrial and medical settings. Existing 4D-CT reconstructions are designed for sparse sampling schemes that require fast CT scanners to capture multiple, rapid revolutions around the scene in order to generate high quality results. However, if the scene is moving too fast, then the sampling occurs along a limited view and is difficult to reconstruct due to spatiotemporal ambiguities. In this work, we design a reconstruction pipeline using implicit neural representations coupled with a novel parametric motion field warping to perform limited view 4D-CT reconstruction of rapidly deforming scenes. Importantly, we utilize a differentiable analysis-by-synthesis approach to compare with captured x-ray sinogram data in a self-supervised fashion. Thus, our resulting optimization method requires no training data to reconstruct the scene. We demonstrate that our proposed system robustly reconstructs scenes containing deformable and periodic motion and validate against state-of-the-art baselines. Further, we demonstrate an ability to reconstruct continuous spatiotemporal representations of our scenes and upsample them to arbitrary volumes and frame rates post-optimization. This research opens a new avenue for implicit neural representations in computed tomography reconstruction in general.

eess.IV↗
arXiv2020

Unsupervised deep learning for super-resolution reconstruction of turbulence

Recent attempts to use deep learning for super-resolution reconstruction of turbulent flows have used supervised learning, which requires paired data for training. This limitation hinders more practical applications of super-resolution reconstruction. Therefore, we present an unsupervised learning model that adopts a cycle-consistent generative adversarial network that can be trained with unpaired turbulence data for super-resolution reconstruction. Our model is validated using three examples: (i) recovering the original flow field from filtered data using direct numerical simulation (DNS) of homogeneous isotropic turbulence; (ii) reconstructing full-resolution fields using partially measured data from the DNS of turbulent channel flows; and (iii) generating a DNS-resolution flow field from large eddy simulation (LES) data for turbulent channel flows. In examples (i) and (ii), for which paired data are available for supervised learning, our unsupervised model demonstrates qualitatively and quantitatively similar performance as that of the best supervised-learning model. More importantly, in example (iii), where supervised learning is impossible, our model successfully reconstructs the high-resolution flow field of statistical DNS quality from the LES data. This demonstrates that unsupervised learning of turbulence data is indeed possible, opening a new door for the wide application of super-resolution reconstruction of turbulent fields.

physics.flu-dyn↗
arXiv2020

Improved Protein-ligand Binding Affinity Prediction with Structure-Based Deep Fusion Inference

Predicting accurate protein-ligand binding affinity is important in drug discovery but remains a challenge even with computationally expensive biophysics-based energy scoring methods and state-of-the-art deep learning approaches. Despite the recent advances in the deep convolutional and graph neural network based approaches, the model performance depends on the input data representation and suffers from distinct limitations. It is natural to combine complementary features and their inference from the individual models for better predictions. We present fusion models to benefit from different feature representations of two neural network models to improve the binding affinity prediction. We demonstrate effectiveness of the proposed approach by performing experiments with the PDBBind 2016 dataset and its docking pose complexes. The results show that the proposed approach improves the overall prediction compared to the individual neural network models with greater computational efficiency than related biophysics based energy scoring functions. We also discuss the benefit of the proposed fusion inference with several example complexes. The software is made available as open source at https://github.com/llnl/fast.

q-bio.BM↗
arXiv2020

Improving Limited Angle CT Reconstruction with a Robust GAN Prior

Limited angle CT reconstruction is an under-determined linear inverse problem that requires appropriate regularization techniques to be solved. In this work we study how pre-trained generative adversarial networks (GANs) can be used to clean noisy, highly artifact laden reconstructions from conventional techniques, by effectively projecting onto the inferred image manifold. In particular, we use a robust version of the popularly used GAN prior for inverse problems, based on a recent technique called corruption mimicking, that significantly improves the reconstruction quality. The proposed approach operates in the image space directly, as a result of which it does not need to be trained or require access to the measurement model, is scanner agnostic, and can work over a wide range of sensing scenarios.

eess.IV↗