Meta tags:
author=Byungsoo Kim;
description=;
keywords=jekyll, jekyll-theme, academic-website, portfolio-website;

Headings (most frequently used words):

byungsoo, kim, news, publications,

Text of the page (most frequently used words):
the (116), and (92), for (42), our (36), that (30), with (29), neural (29), fluid (24), #byungsoo (21), kim (20), data (19), url (18), this (18), method (18), simulations (18), from (18), smoke (18), based (17), more (16), simulation (16), are (15), abs (15), network (14), time (14), flow (14), which (11), networks (11), paper (11), model (11), fields (11), space (11), control (11), transfer (11), vector (10), field (10), deep (10), set (9), proc (9), generative (9), training (9), can (9), latent (9), style (9), resolution (9), 2024 (8), pdf (8), graphics (8), convolutional (8), fluids (8), video (8), vinicius (8), azevedo (8), authors (8), jul (8), stylization (8), high (8), physics (8), learning (7), present (7), images (7), results (7), slides (7), code (7), computer (7), eurographics (7), trained (7), velocity (7), such (7), arxiv (7), input (7), 2020 (7), remote (7), magnetic (7), ice (7), 2023 (7), abstract (6), forum (6), author (6), using (6), proposed (6), extraction (6), end (6), density (6), approach (6), quality (6), navigation (6), facial (6), super (6), work (5), image (5), however (5), semantic (5), demonstrate (5), sketches (5), may (5), parameters (5), different (5), than (5), reference (5), challenging (5), transport (5), volumetric (5), allows (5), acm (5), artist (5), multi (5), spatial (5), modeling (5), scale (5), igm (5), animation (5), sensing (5), urban (5), technology (5), eth (4), zurich (4), line (4), not (4), all (4), these (4), user (4), paths (4), use (4), existing (4), features (4), loss (4), complex (4), models (4), compression (4), feature (4), near (4), two (4), inputs (4), controlling (4), dynamics (4), capture (4), temporal (4), trans (4), graph (4), siggraph (4), between (4), low (4), frequency (4), reconstruction (4), show (4), shape (4), color (4), stokes (4), systems (4), physical (4), test (4), 2022 (4), soft (4), bodies (4), actuation (4), deformation (4), cost (4), lidar (4), research (4), aerodynamic (4), roughness (4), raster (3), art (3), especially (3), therefore (3), segmentation (3), most (3), computed (3), approaches (3), into (3), provide (3), predictions (3), dataset (3), novel (3), reduced (3), due (3), able (3), while (3), function (3), generated (3), 2019 (3), parameterized (3), robust (3), visualization (3), unsteady (3), recently (3), towards (3), steady (3), frames (3), numerically (3), pipeline (3), obtain (3), contribution (3), efficiently (3), artistically (3), has (3), been (3), techniques (3), target (3), often (3), artists (3), structures (3), hence (3), general (3), aware (3), since (3), markus (3), gross (3), bands (3), improves (3), short (3), motion (3), lagrangian (3), enables (3), predict (3), long (3), term (3), evaluate (3), several (3), website (3), methods (3), used (3), nonlinear (3), magnetization (3), performance (3), permits (3), highly (3), efficient (3), differentiable (3), implicit (3), expressions (3), study (3), building (3), exploration (3), real (3), nvidia (3), elemental (3), prof (3), hosted (2), github (2), previous (2), thesis (2), representations (2), vectorize (2), could (2), same (2), useful (2), further (2), desired (2), computing (2), likely (2), have (2), once (2), vectorization (2), each (2), then (2), combine (2), output (2), train (2), full (2), context (2), random (2), synthetic (2), human (2), drawn (2), cases (2), accurately (2), synthesize (2), representative (2), plausible (2), optimized (2), divergence (2), times (2), addition (2), second (2), wide (2), variety (2), thus (2), enabling (2), applications (2), interpolation (2), faster (2), cpu (2), jun (2), directed (2), their (2), computation (2), frame (2), given (2), optimization (2), supervised (2), requires (2), large (2), amount (2), creation (2), any (2), parametric (2), serve (2), architecture (2), unseen (2), task (2), textures (2), limited (2), patterns (2), flows (2), turbulent (2), propose (2), first (2), algorithm (2), manipulations (2), ranging (2), simple (2), inspired (2), source (2), direct (2), over (2), incompressible (2), consistency (2), stylized (2), reconstructions (2), nov (2), functions (2), frequencies (2), higher (2), directly (2), details (2), analyze (2), focus (2), papers (2), manual (2), transfers (2), onto (2), information (2), extend (2), pose (2), visual (2), production (2), particles (2), benefits (2), grid (2), practical (2), representation (2), robustly (2), problems (2), stable (2), sequences (2), linear (2), cnn (2), subdivision (2), quantities (2), parts (2), encoded (2), domain (2), result (2), furthermore (2), driven (2), tools (2), article (2), developed (2), continuous (2), electromagnetic (2), accurate (2), similar (2), current (2), state (2), jan (2), 2021 (2), transactions (2), glacier (2), its (2), evolution (2), glaciers (2), icefields (2), processing (2), when (2), modelling (2), dimensional (2), together (2), friendly (2), dec (2), guillaume (2), pre (2), keyframes (2), volumes (2), realistic (2), non (2), compute (2), prototyping (2), sketch (2), optimize (2), designed (2), computational (2), interactive (2), supp (2), active (2), point (2), simulator (2), informed (2), corrector (2), various (2), ucps (2), area (2), emerging (2), morphometrical (2), framework (2), simulators (2), other (2), where (2), park (2), realtime (2), publications (2), aug (2), share (2), award (2), published (2), science (2), news (2), senior (2), software (2), engineer (2), copyright, powered, theme, pages, folio, jekyll, 2008, 2014, works, 2016, master, structured, geometry, inverting, rasterization, procedure, inherently, ill, conditioned, there, exist, many, possible, yield, equally, performing, edits, define, problem, instance, created, path, final, determine, pair, clues, help, resolve, ambiguities, intersection, overlap, regions, made, considering, globally, combined, solving, markov, mrf, flexibility, generating, character, datasets, composed, system, recovers, adhere, semantics, drawings, 2018, oliver, wang, cengiz, öztireli, drawing, presents, collection, discrete, parameterizable, capability, architectures, learn, approximate, providing, interpolated, betweens, guarantees, free, handle, parameterizations, spaces, advance, integrating, behaviors, fast, construction, sampling, reconstructed, 700, simulating, underlying, solver, achieving, rates, 1300, nils, thuerey, integral, part, scientific, analysis, researchers, attention, vortex, endeavor, utilize, steps, manner, filtering, conditioning, noisy, resampling, artifacts, stabler, typically, benchmark, generally, local, vatistas, profile, formulate, mixture, parameterize, example, thousands, evaluated, numerical, eurovis, tobias, günther, always, rely, approximating, states, configurations, handcrafted, indirectly, patch, synthesis, either, adding, structural, augmenting, coarse, cannot, spectrum, styles, semantically, tnst, natural, content, intricate, motifs, physically, computes, configuration, optimizing, irrotational, potentials, ensured, transporting, aligning, subsequent, velocities, seamlessly, merging, stylizations, camera, viewpoints, asia, were, employed, fully, reconstruct, convolutions, traditionally, discriminate, error, minimized, correlates, perceived, missing, easily, noticeable, specific, during, mid, yielding, perceptually, better, requiring, comparable, simon, biland, presented, transportbased, simplifies, workflows, arbitrary, only, modifies, but, omits, complete, transferring, successfully, colored, consistently, fabienne, christen, appearance, major, challenges, effects, formulated, viewpoint, unique, compared, attributes, stored, trivially, transported, particle, intrinsically, ensures, structure, notably, resulting, simultaneously, expensive, recursive, alignment, unnecessary, reducing, less, hour, rendering, settings, moreover, artistic, consistent, generality, liquids, likewise, stability, single, phase, navier, equations, relevant, range, achieve, execution, combination, prediction, consists, stacked, memory, lstm, layers, core, lss, separate, respective, individual, distinctively, alter, without, interfering, remaining, values, maximizes, external, selectively, overwriting, predicted, points, capable, like, highlight, recurrent, performed, thoroughly, discuss, components, sca, steffen, wiewel, controllable, doctoral, manipulation, untethered, micro, nanorobots, well, tethered, surgical, minimally, invasive, medicine, mathematical, fundamental, biomedical, describe, compare, newly, clinical, interactions, electromagnets, renders, introduce, adapt, correct, ieee, robotics, samuel, charreyron, quentin, boehler, introduces, instructed, simulates, mass, balance, coupling, sheets, novelty, hybrid, sia, ssa, doing, computationally, demanding, component, substituted, cheap, emulator, mountain, 1000, ones, central, units, fidelity, levels, above, terms, solutions, leading, nearly, identical, transient, thickness, switching, gpu, additional, significant, speed, ups, emulating, open, python, deals, gridded, companion, library, emulators, mechanically, journal, glaciology, jouvet, cordonnier, speeds, orders, magnitude, creative, processes, start, hand, illustrating, object, visualizing, applied, materials, authored, must, trivial, remains, consuming, process, bridging, gap, early, stage, initial, volume, estimate, iteratively, updater, sketcher, embedded, augmentation, strategy, applicability, depicting, both, shapes, robustness, authoring, sessions, rapid, ideas, novice, users, xingchang, huang, laura, wuelfroth, densities, affect, through, internal, mechanism, induces, recent, utilizes, quasi, static, layer, signals, key, formulation, defining, mapping, material, value, property, signal, dominant, making, discretization, agnostic, widely, applicable, mandible, kinematics, particular, case, reliably, reproduce, captured, apply, poses, demonstrating, properties, invariance, honorable, mention, lingchen, yang, gaspard, zoss, actuated, notoriously, difficult, fact, cause, unphysical, aims, balancing, deformations, preservation, characteristics, workflow, diverse, carefully, objectives, they, lead, eventually, visually, appealing, modifications, edited, jingwei, tang, proposes, light, detection, support, morphometric, estimating, length, placement, scales, three, cloud, 3dpc, graphical, interface, gui, was, explore, related, canopy, incheon, metropolitan, south, korea, viable, address, differences, resolutions, although, validating, morphological, dense, tall, towers, substitutes, efforts, required, link, measured, form, regulations, floor, ratio, expand, quantitative, qualitative, knowledge, seung, man, chaeyeon, realistically, enhance, produced, level, detail, closely, approximates, off, much, texttimes, element, count, examples, rooted, ability, construct, paired, respectively, correspondence, face, exemplar, creating, congruence, achieved, simply, dialing, muscle, controls, skeletal, generalizes, compensates, discrepancies, cutting, approximations, variant, does, require, descriptors, provided, efficacy, expressive, performances, comparisons, ablation, experiments, variations, alternatives, scheme, available, https, com, hjoonpark, sim, res, git, hyojoon, sangeetha, grama, srinivasan, matthew, cong, joined, team, led, ken, museth, conditionally, accepted, film, pixar, released, technique, check, out, post, mar, posted, about, gave, talk, postech, jisung, hair, neuralvdb, featured, omniverse, demo, thrilled, along, raphael, ortiz, paul, kanyuk, nominated, category, collaborative, ves, excited, selected, recipient, international, congress, basic, frontiers, currently, developing, received, joint, phd, under, supervision, interests, span, machine, barbara, solenthaler, disney, lab, www, toggle,

Text of the page (random words):
paired frames from the low and high resolution simulators respectively that are in semantic correspondence with each other we use face animation as an exemplar of such a simulation domain where creating this semantic congruence is achieved by simply dialing in the same muscle actuation controls and skeletal pose in the two simulators our proposed neural network super resolution framework generalizes from this training set to unseen expressions compensates for modeling discrepancies between the two simulations due to limited resolution or cost cutting approximations in the real time variant and does not require any semantic descriptors or parameters to be provided as input other than the result of the real time simulation we evaluate the efficacy of our pipeline on a variety of expressive performances and provide comparisons and ablation experiments for plausible variations and alternatives to our proposed scheme our code is available at https github com hjoonpark 3d sim super res git study on morphometrical urban aerodynamic roughness multi scale exploration using lidar remote sensing seung man an byungsoo kim chaeyeon yi and 3 more authors remote sensing 2024 abs pdf url this study proposes the use of light detection and ranging lidar remote sensing rs to support morphometric research for estimating the aerodynamic roughness length z0 of building placement on various scales a lidar three dimensional point cloud 3dpc data processing graphical user interface gui was developed to explore the z0 and related urban canopy parameters ucps in the incheon metropolitan area in south korea the results show that multi scale urban aerodynamic roughness exploration is viable and can address differences in urban building data at various spatial resolutions although validating morphological multi scale ucps using dense tall towers is challenging emerging low cost and efficient methods can serve as substitutes however further efforts are required to link the measured z0 to building form regulations such as floor area ratio and expand rs research to obtain more quantitative and qualitative knowledge physics informed neural corrector for deformation based fluid control jingwei tang byungsoo kim vinicius c azevedo and 1 more author computer graphics forum proc eurographics may 2023 abs pdf url video website abstract controlling fluid simulations is notoriously difficult due to its high computational cost and the fact that user control inputs can cause unphysical motion we present an interactive method for deformation based fluid control our method aims at balancing the direct deformations of fluid fields and the preservation of physical characteristics we train convolutional neural networks with physics inspired loss functions together with a differentiable fluid simulator and provide an efficient workflow for flow manipulations at test time we demonstrate diverse test cases to analyze our carefully designed objectives and show that they lead to physical and eventually visually appealing modifications on edited fluid data implicit neural representation for physics driven actuated soft bodies lingchen yang byungsoo kim gaspard zoss and 3 more authors acm trans graph proc siggraph jul 2022 honorable mention abs pdf supp url video website active soft bodies can affect their shape through an internal actuation mechanism that induces a deformation similar to recent work this paper utilizes a differentiable quasi static and physics based simulation layer to optimize for actuation signals parameterized by neural networks our key contribution is a general and implicit formulation to control active soft bodies by defining a function that enables a continuous mapping from a spatial point in the material space to the actuation value this property allows us to capture the signal s dominant frequencies making the method discretization agnostic and widely applicable we extend our implicit model to mandible kinematics for the particular case of facial animation and show that we can reliably reproduce facial expressions captured with high quality capture systems we apply the method to volumetric soft bodies human poses and facial expressions demonstrating artist friendly properties such as simple control over the latent space and resolution invariance at test time deep reconstruction of 3d smoke densities from artist sketches byungsoo kim xingchang huang laura wuelfroth and 4 more authors computer graphics forum proc eurographics may 2022 abs pdf supp code slides url video abstract creative processes of artists often start with hand drawn sketches illustrating an object pre visualizing these keyframes is especially challenging when applied to volumetric materials such as smoke the authored 3d density volumes must capture realistic flow details and turbulent structures which is highly non trivial and remains a manual and time consuming process we therefore present a method to compute a 3d smoke density field directly from 2d artist sketches bridging the gap between early stage prototyping of smoke keyframes and pre visualization from the sketch inputs we compute an initial volume estimate and optimize the density iteratively with an updater cnn our differentiable sketcher is embedded into the end to end training which results in robust reconstructions our training data set and sketch augmentation strategy are designed such that it enables general applicability we evaluate the method on synthetic inputs and sketches from artists depicting both realistic smoke volumes and highly non physical smoke shapes the high computational performance and robustness of our method at test time allows interactive authoring sessions of volumetric density fields for rapid prototyping of ideas by novice users deep learning speeds up ice flow modelling by several orders of magnitude guillaume jouvet guillaume cordonnier byungsoo kim and 3 more authors journal of glaciology dec 2021 abs pdf url this paper introduces the instructed glacier model igm a model that simulates ice dynamics mass balance and its coupling to predict the evolution of glaciers icefields or ice sheets the novelty of igm is that it models the ice flow by a convolutional neural network which is trained from data generated with hybrid sia ssa or stokes ice flow models by doing so the most computationally demanding model component is substituted by a cheap emulator once trained with representative data we demonstrate that igm permits to model mountain glaciers up to 1000 faster than stokes ones on central processing units cpu with fidelity levels above 90 in terms of ice flow solutions leading to nearly identical transient thickness evolution switching to the gpu often permits additional significant speed ups especially when emulating stokes dynamics or and modelling at high spatial resolution igm is an open source python code which deals with two dimensional 2 d gridded input and output data together with a companion library of trained ice flow emulators igm permits user friendly highly efficient and mechanically state of the art glacier and icefields simulations t ro modeling electromagnetic navigation systems samuel l charreyron quentin boehler byungsoo kim and 3 more authors ieee transactions on robotics jan 2021 abs pdf url remote magnetic navigation is used for the manipulation of untethered micro and nanorobots as well as tethered magnetic surgical tools for minimally invasive medicine mathematical modeling of the magnetic fields generated by magnetic navigation systems is a fundamental task in the control of such tools for biomedical applications in this article we describe and compare several existing and newly developed methods for representations of continuous magnetic fields using interpolation in the context of remote magnetic navigation clinical scale electromagnetic navigation systems feature nonlinear magnetization and magnetization interactions between electromagnets which renders accurate magnetic field modeling challenging we first introduce a method that can adapt existing linear models to correct for nonlinear magnetization with similar performance to the current state of the art nonlinear model furthermore we present a method based on convolutional neural networks data driven methods for artist directed fluid simulations byungsoo kim doctoral thesis eth zurich 2020 url latent space subdivision stable and controllable time predictions for fluid flow steffen wiewel byungsoo kim vinicius c azevedo and 2 more authors computer graphics forum proc sca nov 2020 abs arxiv url website abstract we propose an end to end trained neural network architecture to robustly predict the complex dynamics of fluid flows with high temporal stability we focus on single phase smoke simulations in 2d and 3d based on the incompressible navier stokes ns equations which are relevant for a wide range of practical problems to achieve stable predictions for long term flow sequences with linear execution times a convolutional neural network cnn is trained for spatial compression in combination with a temporal prediction network that consists of stacked long short term memory lstm layers our core contribution is a novel latent space subdivision lss to separate the respective input quantities into individual parts of the encoded latent space domain as a result this allows to distinctively alter the encoded quantities without interfering with the remaining latent space values and hence maximizes external control by selectively overwriting parts of the predicted latent space points our proposed method is capable to robustly predict long term sequences of complex physics problems like the flow of fluids in addition we highlight the benefits of a recurrent training on the latent space creation which is performed by the spatial compression network furthermore we thoroughly evaluate and discuss several different components of our method lagrangian neural style transfer for fluids byungsoo kim vinicius c azevedo markus gross and 1 more author acm trans graph proc siggraph jul 2020 abs arxiv code url video artistically controlling the shape motion and appearance of fluid simulations pose major challenges in visual effects production in this paper we present a neural style transfer approach from images to 3d fluids formulated in a lagrangian viewpoint using particles for style transfer has unique benefits compared to grid based techniques attributes are stored on the particles and hence are trivially transported by the particle motion this intrinsically ensures temporal consistency of the optimized stylized structure and notably improves the resulting quality simultaneously the expensive recursive alignment of stylization velocity fields of grid approaches is unnecessary reducing the computation time to less than an hour and rendering neural flow stylization practical in production settings moreover the lagrangian representation improves artistic control as it allows for multi fluid stylization and consistent color transfer from images and the generality of the method enables stylization of smoke and liquids likewise neural smoke stylization with color transfer fabienne christen byungsoo kim vinicius c azevedo and 1 more author in eurographics 2020 short papers may 2020 abs arxiv slides url video artistically controlling fluid simulations requires a large amount of manual work by an artist the recently presented transportbased neural style transfer approach simplifies workflows as it transfers the style of arbitrary input images onto 3d smoke simulations however the method only modifies the shape of the fluid but omits color information in this work we therefore extend the previous approach to obtain a complete pipeline for transferring shape and color information onto 2d and 3d smoke simulations with neural networks our results demonstrate that our method successfully transfers colored style features consistently in space and time to smoke data for different input textures frequency aware reconstruction of fluid simulations with generative networks simon biland vinicius c azevedo byungsoo kim and 1 more author in eurographics 2020 short papers may 2020 abs arxiv slides url convolutional neural networks were recently employed to fully reconstruct fluid simulation data from a set of reduced parameters however since de convolutions traditionally trained with supervised l1 loss functions do not discriminate between low and high frequencies in the data the error is not minimized efficiently for higher bands this directly correlates with the quality of the perceived results since missing high frequency details are easily noticeable in this paper we analyze the reconstruction quality of generative networks and present a frequency aware loss function that is able to focus on specific bands of the dataset during training time we show that our approach improves reconstruction quality of fluid simulation data in mid frequency bands yielding perceptually better results while requiring comparable training time transport based neural style transfer for smoke simulations byungsoo kim vinicius c azevedo markus gross and 1 more author acm trans graph proc siggraph asia nov 2019 abs arxiv code slides url video artistically controlling fluids has always been a challenging task optimization techniques rely on approximating simulation states towards target velocity or density field configurations which are often handcrafted by artists to indirectly control smoke dynamics patch synthesis techniques transfer image textures or simulation features to a target flow field however these are either limited to adding structural patterns or augmenting coarse flows with turbulent structures and hence cannot capture the full spectrum of different styles and semantically complex structures in this paper we propose the first transport based neural style transfer tnst algorithm for volumetric smoke data our method is able to transfer features from natural images to smoke simulations enabling general content aware manipulations ranging from simple patterns to intricate motifs the proposed algorithm is physically inspired since it computes the density transport from a source input smoke to a desired target configuration our transport based approach allows direct control over the divergence of the stylization velocity field by optimizing incompressible and irrotational potentials that transport smoke towards stylization temporal consistency is ensured by transporting and aligning subsequent stylized velocities and 3d reconstructions are computed by seamlessly merging stylizations from different camera viewpoints robust reference frame extraction from unsteady 2d vector fields with convolutional neural networks byungsoo kim and tobias günther computer graphics forum proc eurovis jul 2019 abs arxiv slides url abstract robust feature extraction is an integral part of scientific visualization in unsteady vector field analysis researchers recently directed their ...