arXiv subjects

Taehun Kim

Publications and source records attributed to Taehun Kim.

At least 37 records · Page 2Linked to original sources

arXiv2024

Strict condition for the $L^{2}$-wellposedness of fifth and sixth order dispersive equations

We provide a set of conditions that is necessary and sufficient for the $L^{2}$-wellposedness of the Cauchy problem for fifth and sixth order variable-coefficient linear dispersive equations. The necessity of these conditions had been presented by Tarama, and we scrutinized their proof to split the conditions into several parts so that an inductive argument is applicable. This inductive argument simplifies the engineering process of the appropriate pseudodifferential operator needed for the proof of $L^{2}$-wellposedness.

math.AP↗

arXiv2024

PIM-MMU: A Memory Management Unit for Accelerating Data Transfers in Commercial PIM Systems

Processing-in-memory (PIM) has emerged as a promising solution for accelerating memory-intensive workloads as they provide high memory bandwidth to the processing units. This approach has drawn attention not only from the academic community but also from the industry, leading to the development of real-world commercial PIM devices. In this work, we first conduct an in-depth characterization on UPMEM's general purpose PIM system and analyze the bottlenecks caused by the data transfers across the DRAM and PIM address space. Our characterization study reveals several critical challenges associated with DRAM to/from PIM data transfers in memory bus integrated PIM systems, for instance, its high CPU core utilization, high power consumption, and low read/write throughput for both DRAM and PIM. Driven by our key findings, we introduce the PIM-MMU architecture which is a hardware/software codesign that enables energy-efficient DRAM to/from PIM transfers for PIM systems. PIM-MMU synergistically combines a hardwarebased data copy engine, a PIM-optimized memory scheduler, and a heterogeneity-aware memory mapping function, the utilization of which is supported by our PIM-MMU software stack, significantly improving the efficiency of DRAM to/from PIM data transfers. Experimental results show that PIM-MMU improves the DRAM to/from PIM data transfer throughput by an average 4.1x and enhances its energy-efficiency by 4.1x, leading to a 2.2x end-to-end speedup for real-world PIM workloads.

cs.AR↗

arXiv2024

DeepOnto: A Python Package for Ontology Engineering with Deep Learning

Integrating deep learning techniques, particularly language models (LMs), with knowledge representation techniques like ontologies has raised widespread attention, urging the need of a platform that supports both paradigms. Although packages such as OWL API and Jena offer robust support for basic ontology processing features, they lack the capability to transform various types of information within ontologies into formats suitable for downstream deep learning-based applications. Moreover, widely-used ontology APIs are primarily Java-based while deep learning frameworks like PyTorch and Tensorflow are mainly for Python programming. To address the needs, we present DeepOnto, a Python package designed for ontology engineering with deep learning. The package encompasses a core ontology processing module founded on the widely-recognised and reliable OWL API, encapsulating its fundamental features in a more "Pythonic" manner and extending its capabilities to incorporate other essential components including reasoning, verbalisation, normalisation, taxonomy, projection, and more. Building on this module, DeepOnto offers a suite of tools, resources, and algorithms that support various ontology engineering tasks, such as ontology alignment and completion, by harnessing deep learning methods, primarily pre-trained LMs. In this paper, we also demonstrate the practical utility of DeepOnto through two use-cases: the Digital Health Coaching in Samsung Research UK and the Bio-ML track of the Ontology Alignment Evaluation Initiative (OAEI).

cs.AI↗

arXiv2024

Pathfinding Future PIM Architectures by Demystifying a Commercial PIM Technology

Processing-in-memory (PIM) has been explored for decades by computer architects, yet it has never seen the light of day in real-world products due to their high design overheads and lack of a killer application. With the advent of critical memory-intensive workloads, several commercial PIM technologies have been introduced to the market ranging from domain-specific PIM architectures to more general-purpose PIM architectures. In this work, we deepdive into UPMEM's commercial PIM technology, a general-purpose PIM-enabled parallel architecture that is highly programmable. Our first key contribution is the development of a flexible simulation framework for PIM. The simulator we developed (aka PIMulator) enables the compilation of UPMEM-PIM source codes into its compiled machine-level instructions, which are subsequently consumed by our cycle-level performance simulator. Using PIMulator, we demystify UPMEM's PIM design through a detailed characterization study. Building on top of our characterization, we conduct a series of case studies to pathfind important architectural features that we deem will be critical for future PIM architectures to support

cs.AR↗

arXiv2023

TempGNN: Temporal Graph Neural Networks for Dynamic Session-Based Recommendations

Session-based recommendations which predict the next action by understanding a user's interaction behavior with items within a relatively short ongoing session have recently gained increasing popularity. Previous research has focused on capturing the dynamics of sequential dependencies from complicated item transitions in a session by means of recurrent neural networks, self-attention models, and recently, mostly graph neural networks. Despite the plethora of different models relying on the order of items in a session, few approaches have been proposed for dealing better with the temporal implications between interactions. We present Temporal Graph Neural Networks (TempGNN), a generic framework for capturing the structural and temporal dynamics in complex item transitions utilizing temporal embedding operators on nodes and edges on dynamic session graphs, represented as sequences of timed events. Extensive experimental results show the effectiveness and adaptability of the proposed method by plugging it into existing state-of-the-art models. Finally, TempGNN achieved state-of-the-art performance on two real-world e-commerce datasets.

cs.IR↗

arXiv2023

Bond-dependent anisotropy and magnon decay in cobalt-based Kitaev triangular antiferromagnet

The Kitaev model, a honeycomb network of spins with bond-dependent anisotropic interactions, is a rare example of having a quantum spin liquid ground state. Although most Kitaev model candidate materials eventually order magnetically due to inevitable non-Kitaev terms, their bond-dependent anisotropy manifests in unusual spin dynamics. It has recently been suggested that bond-dependent anisotropy can stabilise novel magnetic phases and exotic spin dynamics on the geometrically frustrated triangular lattice. However, few materials have been identified with simultaneous geometric frustration and bond-dependent anisotropy. Here, we report a frustrated triangular lattice with bond-dependent anisotropy in the cobalt-based triangular van der Waals antiferromagnet CoI2. Its momentum and energy-resolved spin dynamics exhibit substantial magnon breakdown and complex level repulsion, as measured by inelastic neutron scattering. A thorough examination of excitations in both the paramagnetic and magnetically ordered states reveals that the bond-dependent anisotropy is the origin of the spiral order and the magnon breakdown found in CoI2. Our result paves the way toward a new research direction for the Kitaev model with geometrical frustration.

cond-mat.str-el↗

arXiv2023

Single- and Multimagnon Dynamics in Antiferromagnetic $α$-Fe$_2$O$_3$ Thin Films

Understanding the spin dynamics in antiferromagnetic (AFM) thin films is fundamental for designing novel devices based on AFM magnon transport. Here, we study the magnon dynamics in thin films of AFM $S=5/2$ $α$-Fe$_2$O$_3$ by combining resonant inelastic x-ray scattering, Anderson impurity model plus dynamical mean-field theory, and Heisenberg spin model. Below 100 meV, we observe the thickness-independent (down to 15 nm) acoustic single-magnon mode. At higher energies (100-500 meV), an unexpected sequence of equally spaced, optical modes is resolved and ascribed to $ΔS_z = 1$, 2, 3, 4, and 5 magnetic excitations corresponding to multiple, noninteracting magnons. Our study unveils the energy, character, and momentum-dependence of single and multimagnons in $α$-Fe$_2$O$_3$ thin films, with impact on AFM magnon transport and its related phenomena. From a broader perspective, we generalize the use of L-edge resonant inelastic x-ray scattering as a multispin-excitation probe up to $ΔS_z = 2S$. Our analysis identifies the spin-orbital mixing in the valence shell as the key element for accessing excitations beyond $ΔS_z = 1$, and up to, e.g., $ΔS_z = 5$. At the same time, we elucidate the novel origin of the spin excitations beyond the $ΔS_z = 2$, emphasizing the key role played by the crystal lattice as a reservoir of angular momentum that complements the quanta carried by the absorbed and emitted photons.

cond-mat.str-el↗

arXiv2022

Revisiting Image Pyramid Structure for High Resolution Salient Object Detection

Salient object detection (SOD) has been in the spotlight recently, yet has been studied less for high-resolution (HR) images. Unfortunately, HR images and their pixel-level annotations are certainly more labor-intensive and time-consuming compared to low-resolution (LR) images and annotations. Therefore, we propose an image pyramid-based SOD framework, Inverse Saliency Pyramid Reconstruction Network (InSPyReNet), for HR prediction without any of HR datasets. We design InSPyReNet to produce a strict image pyramid structure of saliency map, which enables to ensemble multiple results with pyramid-based image blending. For HR prediction, we design a pyramid blending method which synthesizes two different image pyramids from a pair of LR and HR scale from the same image to overcome effective receptive field (ERF) discrepancy. Our extensive evaluations on public LR and HR SOD benchmarks demonstrate that InSPyReNet surpasses the State-of-the-Art (SotA) methods on various SOD metrics and boundary accuracy.

cs.CV↗

arXiv2022

Accurate Bundle Matching and Generation via Multitask Learning with Partially Shared Parameters

How can we recommend existing bundles to users accurately? How can we generate new tailored bundles for users? Recommending a bundle, or a group of various items, has attracted widespread attention in e-commerce owing to the increased satisfaction of both users and providers. Bundle matching and bundle generation are two representative tasks in bundle recommendation. The bundle matching task is to correctly match existing bundles to users while the bundle generation is to generate new bundles that users would prefer. Although many recent works have developed bundle recommendation models, they fail to achieve high accuracy since they do not handle heterogeneous data effectively and do not learn a method for customized bundle generation. In this paper, we propose BundleMage, an accurate approach for bundle matching and generation. BundleMage effectively mixes user preferences of items and bundles using an adaptive gate technique to achieve high accuracy for the bundle matching. BundleMage also generates a personalized bundle by learning a generation module that exploits a user preference and the characteristic of a given incomplete bundle to be completed. BundleMage further improves its performance using multi-task learning with partially shared parameters. Through extensive experiments, we show that BundleMage achieves up to 6.6% higher nDCG in bundle matching and 6.3x higher nDCG in bundle generation than the best competitors. We also provide qualitative analysis that BundleMage effectively generates bundles considering both the tastes of users and the characteristics of target bundles.

cs.IR↗

arXiv2022

SR-GCL: Session-Based Recommendation with Global Context Enhanced Augmentation in Contrastive Learning

Session-based recommendations aim to predict the next behavior of users based on ongoing sessions. The previous works have been modeling the session as a variable-length of a sequence of items and learning the representation of both individual items and the aggregated session. Recent research has applied graph neural networks with an attention mechanism to capture complicated item transitions and dependencies by modeling the sessions into graph-structured data. However, they still face fundamental challenges in terms of data and learning methodology such as sparse supervision signals and noisy interactions in sessions, leading to sub-optimal performance. In this paper, we propose SR-GCL, a novel contrastive learning framework for a session-based recommendation. As a crucial component of contrastive learning, we propose two global context enhanced data augmentation methods while maintaining the semantics of the original session. The extensive experiment results on two real-world E-commerce datasets demonstrate the superiority of SR-GCL as compared to other state-of-the-art methods.

cs.IR↗

arXiv2022

STING: Self-attention based Time-series Imputation Networks using GAN

Time series data are ubiquitous in real-world applications. However, one of the most common problems is that the time series data could have missing values by the inherent nature of the data collection process. So imputing missing values from multivariate (correlated) time series data is imperative to improve a prediction performance while making an accurate data-driven decision. Conventional works for imputation simply delete missing values or fill them based on mean/zero. Although recent works based on deep neural networks have shown remarkable results, they still have a limitation to capture the complex generation process of the multivariate time series. In this paper, we propose a novel imputation method for multivariate time series data, called STING (Self-attention based Time-series Imputation Networks using GAN). We take advantage of generative adversarial networks and bidirectional recurrent neural networks to learn latent representations of the time series. In addition, we introduce a novel attention mechanism to capture the weighted correlations of the whole sequence and avoid potential bias brought by unrelated ones. Experimental results on three real-world datasets demonstrate that STING outperforms the existing state-of-the-art methods in terms of imputation accuracy as well as downstream tasks with the imputed values therein.

cs.LG↗

arXiv2022

A Style-aware Discriminator for Controllable Image Translation

Current image-to-image translations do not control the output domain beyond the classes used during training, nor do they interpolate between different domains well, leading to implausible results. This limitation largely arises because labels do not consider the semantic distance. To mitigate such problems, we propose a style-aware discriminator that acts as a critic as well as a style encoder to provide conditions. The style-aware discriminator learns a controllable style space using prototype-based self-supervised learning and simultaneously guides the generator. Experiments on multiple datasets verify that the proposed model outperforms current state-of-the-art image-to-image translation methods. In contrast with current methods, the proposed approach supports various applications, including style interpolation, content transplantation, and local image translation.

cs.CV↗

arXiv2021

Slow oxidation of magnetite nanoparticles elucidates the limits of the Verwey transition

Magnetite (Fe3O4) is of fundamental importance as the original magnetic material and also for the Verwey transition near T_V = 125 K, below which a complex lattice distortion and electron orders occur. The Verwey transition is suppressed by strain or chemical doping effects giving rise to well-documented first and second-order regimes, but the origin of the order change is unclear. Here, we show that slow oxidation of monodisperse Fe3O4 nanoparticles leads to an intriguing variation of the Verwey transition that elucidates the doping effects. Exposure to various fixed oxygen pressures at ambient temperature leads to an initial drop to TV minima as low as 70 K after 45-75 days, followed by recovery to a constant value of 95 K after 160 days that persists in all experiments for aging times up to 1070 days. A physical model based on both doping and doping-gradient effects accounts quantitatively for this evolution and demonstrates that the persistent 95 K value corresponds to the lower limit for homogenously doped magnetite and hence for the first order regime. In comparison, further suppression down to 70 K results from inhomogeneous strains that characterize the second-order region. This work demonstrates that slow reactions of nanoparticles can give exquisite control and separation of homogenous and inhomogeneous doping or strain effects on an nm scale and offers opportunities for similar insights into complex electronic and magnetic phase transitions in other materials.

cond-mat.mtrl-sci↗

arXiv2021

Finding essential parts of the brain in rs-fMRI can improve diagnosing ADHD by Deep Learning

Attention Deficit\Hyperactivity Disorder(ADHD) is considered a very common psychiatric disorder, but it is difficult to establish an accurate diagnostic method for ADHD. Recently, with the development of computing resources and machine learning methods, studies have been conducted to classify ADHD using resting-state functional magnetic resonance(rsfMRI) imaging data. However, most of them utilized all areas of the brain for training the models. In this study, as a different way from this approach, we conducted a study to classify ADHD by selecting areas that are essential for using a deep learning model. For the experiment, rsfMRI data provided by ADHD 200 global competition was used. To obtain an integrated result from the multiple sites, each region of the brain was evaluated with Leave one site out cross-validation. As a result, when we only used 15 important region of interest(ROIs) for training, an accuracy of 70.6% was obtained, significantly exceeding the existing results of 68.6% from all ROIs. In addition, to explore the new structure based on SCCNN-RNN, we performed the same experiment with three modified models: (1) Separate Channel CNN RNN with Attention (ASCRNN), (2) Separate Channel dilate CNN RNN with Attention (ASDRNN), (3) Separate Channel CNN slicing RNN with Attention (ASSRNN). As a result, the ASSRNN model provided a high accuracy of 70.46% when training with only 13 important region of interest (ROI). These results show that finding and using the crucial parts of the brain in diagnosing ADHD by Deep Learning can get better results than using all areas.

eess.IV↗

arXiv2021

UACANet: Uncertainty Augmented Context Attention for Polyp Segmentation

We propose Uncertainty Augmented Context Attention network (UACANet) for polyp segmentation which consider a uncertain area of the saliency map. We construct a modified version of U-Net shape network with additional encoder and decoder and compute a saliency map in each bottom-up stream prediction module and propagate to the next prediction module. In each prediction module, previously predicted saliency map is utilized to compute foreground, background and uncertain area map and we aggregate the feature map with three area maps for each representation. Then we compute the relation between each representation and each pixel in the feature map. We conduct experiments on five popular polyp segmentation benchmarks, Kvasir, CVC-ClinicDB, ETIS, CVC-ColonDB and CVC-300, and achieve state-of-the-art performance. Especially, we achieve 76.6% mean Dice on ETIS dataset which is 13.8% improvement compared to the previous state-of-the-art method. Source code is publicly available at https://github.com/plemeri/UACANet

cs.CV↗

arXiv2021

SpaceMeshLab: Spatial Context Memoization and Meshgrid Atrous Convolution Consensus for Semantic Segmentation

Semantic segmentation networks adopt transfer learning from image classification networks which occurs a shortage of spatial context information. For this reason, we propose Spatial Context Memoization (SpaM), a bypassing branch for spatial context by retaining the input dimension and constantly communicating its spatial context and rich semantic information mutually with the backbone network. Multi-scale context information for semantic segmentation is crucial for dealing with diverse sizes and shapes of target objects in the given scene. Conventional multi-scale context scheme adopts multiple effective receptive fields by multiple dilation rates or pooling operations, but often suffer from misalignment problem with respect to the target pixel. To this end, we propose Meshgrid Atrous Convolution Consensus (MetroCon^2) which brings multi-scale scheme into fine-grained multi-scale object context using convolutions with meshgrid-like scattered dilation rates. SpaceMeshLab (ResNet-101 + SpaM + MetroCon^2) achieves 82.0% mIoU in Cityscapes test and 53.5% mIoU on Pascal-Context validation set.

cs.CV↗

arXiv2021

Spin texture induced by non-magnetic doping and spin dynamics in 2D triangular lattice antiferromagnet h-Y(Mn,Al)O3

Novel effects induced by nonmagnetic impurities in frustrated magnets and quantum spin liquid represent a highly nontrivial and interesting problem. A theoretical proposal of extended modulated spin structures induced by doping of such magnets, distinct from the well-known skyrmions has attracted significant interest. Here, we demonstrate that nonmagnetic impurities can produce such extended spin structures in h-YMnO3, a triangular antiferromagnet with noncollinear magnetic order. Using inelastic neutron scattering (INS), we measured the full dynamical structure factor in Al-doped h-YMnO3 and confirmed the presence of magnon damping with a clear momentum dependence. Our theoretical calculations can reproduce the key features of the INS data, supporting the formation of the proposed spin textures. As such, our study provides the first experimental confirmation of the impurity-induced spin textures. It offers new insights and understanding of the impurity effects in a broad class of noncollinear magnetic systems.

cond-mat.str-el↗

arXiv2021

Fast and Accurate Transferability Measurement for Heterogeneous Multivariate Data

Given a set of heterogeneous source datasets with their classifiers, how can we quickly find the most useful source dataset for a specific target task? We address the problem of measuring transferability between source and target datasets, where the source and the target have different feature spaces and distributions. We propose Transmeter, a fast and accurate method to estimate the transferability of two heterogeneous multivariate datasets. We address three challenges in measuring transferability between two heterogeneous multivariate datasets: reducing time, minimizing domain gap, and extracting meaningful homogeneous representations. To overcome the above issues, we utilize a pre-trained source model, an adversarial network, and an encoder-decoder architecture. Extensive experiments on heterogeneous multivariate datasets show that Transmeter gives the most accurate transferability measurement with up to 10.3 times faster performance than its competitor. We also show that selecting the best source data with Transmeter followed by a full transfer leads to the best transfer accuracy and the fastest running time.

cs.LG↗