arXiv subjects

Daehyun Kim

Publications and source records attributed to Daehyun Kim.

At least 19 recordsLinked to original sources

arXiv2026

TC-Next: Zero-Shot Multimodal Cyclone Forecasting

We present TropicalCycloneNext (TC-Next), a multimodal deep learning model that forecasts tropical cyclone track and intensity at $6$-$24$ h leads by leveraging a foundation model's forecast fields of atmospheric kinematic and thermodynamic fields and GridSat infrared satellite imagery. Trained only on GraphCast forecasts over the Western Pacific (WP), yet reliant only on generic atmospheric variables, TC-Next on GraphCast lowers track error by $15$-$44\%$ and intensity error by a factor of $3$-$6$ relative to a conventional, rule-based tracker, TempestExtremes; applied without retraining to the forecast fields of Pangu-Weather and IFS HRES, it stays ahead of TempestExtremes on both. Applied zero-shot to the generic weather fields of WeatherNext Cyclones on the 2025 WP season, TC-Next attains lower intensity error at every lead time, and lower or comparable track error, compared to that model's specialized direct tracker in a deterministic comparison. Our ablation studies show that our multimodal model is able to utilize the additional modality to improve performance in tracking errors at every lead time and in intensity prediction at longer lead times.

cs.LG↗

arXiv2026

UCMNet: Uncertainty-Aware Context Memory Network for Under-Display Camera Image Restoration

Under-display cameras (UDCs) allow for full-screen designs by positioning the imaging sensor underneath the display. Nonetheless, light diffraction and scattering through the various display layers result in spatially varying and complex degradations, which significantly reduce high-frequency details. Current PSF-based physical modeling techniques and frequency-separation networks are effective at reconstructing low-frequency structures and maintaining overall color consistency. However, they still face challenges in recovering fine details when dealing with complex, spatially varying degradation. To solve this problem, we propose a lightweight \textbf{U}ncertainty-aware \textbf{C}ontext-\textbf{M}emory \textbf{Network} (\textbf{UCMNet}), for UDC image restoration. Unlike previous methods that apply uniform restoration, UCMNet performs uncertainty-aware adaptive processing to restore high-frequency details in regions with varying degradations. The estimated uncertainty maps, learned through an uncertainty-driven loss, quantify spatial uncertainty induced by diffraction and scattering, and guide the Memory Bank to retrieve region-adaptive context from the Context Bank. This process enables effective modeling of the non-uniform degradation characteristics inherent to UDC imaging. Leveraging this uncertainty as a prior, UCMNet achieves state-of-the-art performance on multiple benchmarks with 30\% fewer parameters than previous models. Project page: \href{https://kdhrick2222.github.io/projects/UCMNet/}{https://kdhrick2222.github.io/projects/UCMNet}.

cs.CV↗

arXiv2025

Asymptotic Value of Monitoring Structures in Stochastic Games

This paper studies how improved monitoring affects the limit equilibrium payoff set for stochastic games with imperfect public monitoring. We introduce a simple generalization of Blackwell garbling called weighted garbling in order to compare different monitoring structures for this class of games. Our main result is the monotonicity of the limit perfect public equilibrium (PPE) payoff set with respect to this information order. We show that the limit PPE payoff set expands when the monitoring structure gets more informative with respect to the weighted garbling order. We also show that a similar monotonicity holds for strongly symmetric equilibrium for symmetric stochastic games. Finally, we show that our weighted garbling order is useful to compare the limit PPE payoff set for different state transition laws and monitoring structures when the limit feasible payoff set is the same.

econ.TH↗

arXiv2025

FFT-based Selection and Optimization of Statistics for Robust Recognition of Severely Corrupted Images

Improving model robustness in case of corrupted images is among the key challenges to enable robust vision systems on smart devices, such as robotic agents. Particularly, robust test-time performance is imperative for most of the applications. This paper presents a novel approach to improve robustness of any classification model, especially on severely corrupted images. Our method (FROST) employs high-frequency features to detect input image corruption type, and select layer-wise feature normalization statistics. FROST provides the state-of-the-art results for different models and datasets, outperforming competitors on ImageNet-C by up to 37.1% relative gain, improving baseline of 40.9% mCE on severe corruptions.

cs.CV↗

arXiv2025

One-dimensional modeling of blood flow: A comprehensive yet concise review

One-dimensional (1D) blood flow simulations are extensively used in cardiovascular research due to their computational efficiency and effectiveness in analyzing pulse wave dynamics. Despite their versatility and simplicity, there is a lack of a unified, step-by-step guide integrating theoretical derivations with practical implementation details. In this work, we summarize key components for comprehensive 1D blood flow simulations, including the derivation of reduced-order governing equations, the method of characteristics (Riemann invariants), a finite volume-based numerical scheme, boundary conditions (application of Riemann invariants for reflective/non-reflective and 3-element Windkessel outlet boundaries), junction treatments, verification of presented methodologies, and relevant practical applications. Additionally, we provide detailed step-by-step instructions for implementing the numerical scheme, applying boundary conditions, and treatment of junctions. By integrating rigorous theory with practical guidance for implementation, we seek to improve accessibility of 1D blood flow simulations. We anticipate that this guide will serve as a valuable resource and foundational reference for both novice and experienced researchers in cardiovascular modeling.

q-bio.QM↗

arXiv2024

Lightning declines over shipping lanes following regulation of fuel sulfur emissions

Aerosol interactions with clouds represent a significant uncertainty in our understanding of the Earth system. Deep convective clouds may respond to aerosol perturbations in several ways that have proven difficult to elucidate with observations. Here, we leverage the two busiest maritime shipping lanes in the world, which emit aerosol particles and their precursors into an otherwise relatively clean tropical marine boundary layer, to make headway on the influence of aerosol on deep convective clouds. The recent seven-fold change in allowable fuel sulfur by the International Maritime Organization allows us to test the sensitivity of the lightning to changes in ship plume aerosol size distributions. We find that, across a range of atmospheric thermodynamic conditions, the previously documented enhancement of lightning over the shipping lanes has fallen by over 40%. The enhancement is therefore at least partially aerosol-mediated, a conclusion that is supported by observations of droplet number at cloud base, which show a similar decline over the shipping lane. These results have fundamental implications for our understanding of aerosol-cloud interactions, suggesting that deep convective clouds are impacted by the aerosol number distribution in the remote marine environment.

physics.ao-ph↗

arXiv2024

Characterizing the Feasible Payoff Set of OLG Repeated Games

We study the set of (stationary) feasible payoffs of overlapping generation repeated games that can be achieved by action sequences in which every generation of players plays the same sequence of action profiles. First, we completely characterize the set of feasible payoffs given any fixed discount factor of players and the length of interaction. This allows us to obtain the feasible payoff set in closed form. Second, we provide novel comparative statics of the feasible payoff set with respect to the discount factor and the length of interaction. Interestingly, the feasible payoff set becomes smaller as players' discount factor becomes larger. Additionally, we identify a necessary and sufficient condition for this monotonicity to be strict.

econ.TH↗

arXiv2024

DBenVis: A Visual Analytics System for Comparing DBMS Performance via Benchmark Programs

Database benchmarking is an essential method for evaluating and comparing the performance characteristics of a database management system (DBMS). It helps researchers and developers to evaluate the efficacy of their optimizations or newly developed DBMS solutions. Also, companies can benefit by analyzing the performance of DBMS under specific workloads and leveraging the result to select the most suitable system for their needs. The proper interpretation of raw benchmark results requires effective visualization, which helps users gain meaningful insights. However, visualization of the results requires prior knowledge, and existing approaches often involve time-consuming manual tasks. This is due to the absence of a unified visual analytics system for benchmark results across diverse DBMSs. To address these challenges, we present DBenVis, an interactive visual analytics system that provides efficient and versatile benchmark results visualization. DBenVis is designed to support both online transaction processing (OLTP) and online analytic processing (OLAP) benchmarks. DBenVis provides an interactive comparison view, which enables users to perform an in-depth analysis of performance characteristics across various metrics among different DBMSs. Notably, we devise an interactive visual encoding idiom for the OLAP benchmark to represent a query execution plan as a tree. In the process of building a system, we propose novel techniques for parsing meaningful data from raw benchmark results and converting the query plan to a D3 hierarchical format. Through case studies conducted with domain experts, we demonstrate the efficacy and usability of DBenVis.

cs.HC↗

arXiv2024

Weighted Garbling

We introduce an information order on experiments based on weighted garbling, a generalization of the standard notion of garbling. In this order, an experiment is more informative than another if the latter is a weighted garbling of the former. We show that this is equivalent to ordinary garbling conditional on a payoff-irrelevant event. We also characterize the order in terms of induced posterior belief distributions, showing that it depends only on their support. Our main results provide two decision-theoretic characterizations of this order. First, in static decision problems, one experiment dominates another if and only if its value of information is at least a fixed fraction of the other's across all problems. Second, in a class of stopping time problems with a hidden Markov process and repeated experimentation, one experiment dominates another if and only if it yields weakly higher expected payoffs for every problem with a regular prior.

econ.TH↗

arXiv2024

Real-time Digital RF Emulation -- II: A Near Memory Custom Accelerator

A near memory hardware accelerator, based on a novel direct path computational model, for real-time emulation of radio frequency systems is demonstrated. Our evaluation of hardware performance uses both application-specific integrated circuits (ASIC) and field programmable gate arrays (FPGA) methodologies: 1). The ASIC testchip implementation, using TSMC 28nm CMOS, leverages distributed autonomous control to extract concurrency in compute as well as low latency. It achieves a $518$ MHz per channel bandwidth in a prototype $4$-node system. The maximum emulation range supported in this paradigm is $9.5$ km with $0.24$ $μ$s of per-sample emulation latency. 2). The FPGA-based implementation, evaluated on a Xilinx ZCU104 board, demonstrates a $9$-node test case (two Transmitters, one Receiver, and $6$ passive reflectors) with an emulation range of $1.13$ km to $27.3$ km at $215$ MHz bandwidth.

eess.SP↗

arXiv2024

Object-conditioned Bag of Instances for Few-Shot Personalized Instance Recognition

Nowadays, users demand for increased personalization of vision systems to localize and identify personal instances of objects (e.g., my dog rather than dog) from a few-shot dataset only. Despite outstanding results of deep networks on classical label-abundant benchmarks (e.g., those of the latest YOLOv8 model for standard object detection), they struggle to maintain within-class variability to represent different instances rather than object categories only. We construct an Object-conditioned Bag of Instances (OBoI) based on multi-order statistics of extracted features, where generic object detection models are extended to search and identify personal instances from the OBoI's metric space, without need for backpropagation. By relying on multi-order statistics, OBoI achieves consistent superior accuracy in distinguishing different instances. In the results, we achieve 77.1% personal object recognition accuracy in case of 18 personal instances, showing about 12% relative gain over the state of the art.

cs.CV↗

arXiv2024

Deep Neural Network Models Trained With A Fixed Random Classifier Transfer Better Across Domains

The recently discovered Neural collapse (NC) phenomenon states that the last-layer weights of Deep Neural Networks (DNN), converge to the so-called Equiangular Tight Frame (ETF) simplex, at the terminal phase of their training. This ETF geometry is equivalent to vanishing within-class variability of the last layer activations. Inspired by NC properties, we explore in this paper the transferability of DNN models trained with their last layer weight fixed according to ETF. This enforces class separation by eliminating class covariance information, effectively providing implicit regularization. We show that DNN models trained with such a fixed classifier significantly improve transfer performance, particularly on out-of-domain datasets. On a broad range of fine-grained image classification datasets, our approach outperforms i) baseline methods that do not perform any covariance regularization (up to 22%), as well as ii) methods that explicitly whiten covariance of activations throughout training (up to 19%). Our findings suggest that DNNs trained with fixed ETF classifiers offer a powerful mechanism for improving transfer learning across domains.

cs.LG↗

arXiv2020

Coherent Power Scaling in Photonic Crystal Surface Emitting Laser Arrays

A key benefit of photonic crystal surface emitting lasers (PCSELs) is the abillity to increase output power through scaling the emission area while mainting high quality single mode emission, allowing them to close the brightness gap which exists between semiconductor lasers and gas and fibre lasers. However, there are practical limits to the size, and hence power, of an individual PCSEL device and there are trade-offs between single-mode stability and parasitic in-plane losses with increasing device size. In this paper we discuss 2D coherent arrays as an approach to area and coherent power scaling of PCSELs. We demonstrate in two and three element PCSEL arrays an increase in the differential efficiency of the system due to a reduction in in-plane loss.

physics.optics↗

arXiv2020

Iterative Compression of End-to-End ASR Model using AutoML

Increasing demand for on-device Automatic Speech Recognition (ASR) systems has resulted in renewed interests in developing automatic model compression techniques. Past research have shown that AutoML-based Low Rank Factorization (LRF) technique, when applied to an end-to-end Encoder-Attention-Decoder style ASR model, can achieve a speedup of up to 3.7x, outperforming laborious manual rank-selection approaches. However, we show that current AutoML-based search techniques only work up to a certain compression level, beyond which they fail to produce compressed models with acceptable word error rates (WER). In this work, we propose an iterative AutoML-based LRF approach that achieves over 5x compression without degrading the WER, thereby advancing the state-of-the-art in ASR compression.

cs.LG↗

arXiv2020

Attention based on-device streaming speech recognition with large speech corpus

In this paper, we present a new on-device automatic speech recognition (ASR) system based on monotonic chunk-wise attention (MoChA) models trained with large (> 10K hours) corpus. We attained around 90% of a word recognition rate for general domain mainly by using joint training of connectionist temporal classifier (CTC) and cross entropy (CE) losses, minimum word error rate (MWER) training, layer-wise pre-training and data augmentation methods. In addition, we compressed our models by more than 3.4 times smaller using an iterative hyper low-rank approximation (LRA) method while minimizing the degradation in recognition accuracy. The memory footprint was further reduced with 8-bit quantization to bring down the final model size to lower than 39 MB. For on-demand adaptation, we fused the MoChA models with statistical n-gram models, and we could achieve a relatively 36% improvement on average in word error rate (WER) for target domains including the general domain.

eess.AS↗

arXiv2019

ScieNet: Deep Learning with Spike-assisted Contextual Information Extraction

Deep neural networks (DNNs) provide high image classification accuracy, but experience significant performance degradation when perturbation from various sources are present in the input. The lack of resilience to input perturbations makes DNN less reliable for systems interacting with physical world such as autonomous vehicles, robotics, to name a few, where imperfect input is the normal condition. We present a hybrid deep network architecture with spike-assisted contextual information extraction (ScieNet). ScieNet integrates unsupervised learning using spiking neural network (SNN) for unsupervised contextual informationextraction with a back-end DNN trained for classification. The integrated network demonstrates high resilience to input perturbations without relying on prior training on perturbed inputs. We demonstrate ScieNet with different back-end DNNs for image classification using CIFAR dataset considering stochastic (noise) and structured (rain) input perturbations. Experimental results demonstrate significant improvement in accuracy on noisy and rainy images without prior training, while maintaining state-of-the-art accuracy on clean images.

cs.LG↗

arXiv2017

Comments on Nonrenewal Statistics in the Catalytic Activity of Enzyme Molecules at Mesoscopic Concentrations

It is well known in enzyme kinetics that the Michaelis-Menten (MM) equation is applicable only to enzymes in the steady state. We show that the result obtained in the previous work [Phys. Rev. Lett. 107, 218301 (2011)] is inconsistent with the MM equation, not because the authors considered the enzyme system at mesoscopic concentrations but because they considered the enzyme system in the non-stationary state. The substrate concentration dependence of the mean turnover time is, in fact, consistent with the MM equation in the steady state, regardless of the number of enzymes in the system.

physics.bio-ph↗

arXiv2015

Feasibility Study of Neutron Dose for Real Time Image Guided Proton Therapy: A Monte Carlo Study

Two full rotating gantry with different nozzles (Multipurpose nozzle with MLC, Scanning Dedicated nozzle) with conventional cyclotron system is installed and under commissioning for various proton treatment options at Samsung Medical Center in Korea. The purpose of this study is to investigate neutron dose equivalent per therapeutic dose, H/D, to x-ray imaging equipment under various treatment conditions with monte carlo simulation. At first, we investigated H/D with the various modifications of the beam line devices (Scattering, Scanning, Multi-leaf collimator, Aperture, Compensator) at isocenter, 20, 40, 60 cm distance from isocenter and compared with other research groups. Next, we investigated the neutron dose at x-ray equipments used for real time imaging with various treatment conditions. Our investigation showed the 0.07 ~ 0.19 mSv/Gy at x-ray imaging equipments according to various treatment options and intestingly 50% neutron dose reduction effect of flat panel detector was observed due to multi- leaf collimator during proton scanning treatment with multipurpose nozzle. In future studies, we plan to investigate experimental measurement of neutron dose and validation of simulation data for x-ray imaging equipment with additional neutron dose reduction method.

physics.med-ph↗