Meta tags:

Headings (most frequently used words):

news, education, experience, publications, personal, joonghyuk, shin,

Text of the page (most frequently used words):
and (25), 2026 (16), for (16), with (12), #joonghyuk (11), models (10), 2025 (10), video (9), paper (9), jaesik (9), park (9), shin (9), image (8), from (7), project (7), page (7), that (6), code (6), based (6), world (6), motionstream (6), jun (6), yan (6), research (6), generative (5), interactive (5), editing (5), zhu (5), more (4), 2023 (4), present (4), synthesis (4), generation (4), motion (4), single (4), xun (4), huang (4), flow (4), intern (4), feb (4), accepted (4), june (3), also (3), two (3), method (3), long (3), fast (3), drag (3), real (3), dit (3), introduce (3), working (3), iclr (3), eli (3), shechtman (3), human (3), eccv (3), text2tactilegraphics (3), tactile (3), worked (3), lab (3), university (3), stealth (3), nvidia (3), snu (3), last (2), baseball (2), postech (2), team (2), studiogan (2), gans (2), benchmarks (2), tasks (2), minguk (2), kang (2), arxiv (2), tailed (2), using (2), data (2), 2024 (2), instantdrag (2), free (2), improving (2), iccv (2), prompt (2), diverse (2), kim (2), exploring (2), time (2), system (2), controls (2), content (2), oral (2), richard (2), zhang (2), zhengqi (2), drpose (2), direct (2), reward (2), fine (2), tuning (2), poses (2), the (2), matching (2), audio (2), jam (2), graphics (2), wenzhen (2), yuan (2), ava (2), pun (2), ruihan (2), gao (2), text (2), equal (2), sep (2), scientist (2), san (2), francisco (2), building (2), adobe (2), student (2), robotics (2), institute (2), seoul (2), stateful (2), morpheus (2), santa (2), clara (2), spatial (2), intelligence (2), cse (2), will (2), summer (2), visit (2), cmu (2), startup (2), papers (2), can (2), updated, face, looker, big, fan, played, years, captain, catcher, love, animals, live, dog, named, poby, like, pokemon, travelling, fifa, games, tachyons, personal, tpami, comprehensive, library, reproduces, over, popular, providing, extensive, fair, evaluation, protocol, 3500, taxonomy, benchmark, propose, stage, recognition, textual, inverted, tokens, synthesize, images, achieving, sota, results, standard, when, trained, scratch, fill, balancing, siggraph, asia, optimization, pipeline, requires, only, instruction, input, learning, datasets, 230, daehyeon, choi, interactivity, alchan, hwang, perform, systematic, analysis, bidirectional, attention, mechanism, robust, across, sd3, series, flux, daneul, yujin, multimodal, diffusion, transformers, enhanced, streaming, causal, duration, operating, fps, h100, gpu, unlocking, new, possibilities, post, minwoo, huh, improves, pose, accuracy, view, reconstruction, especially, dynamic, acrobatic, without, requiring, expensive, assets, seunguk, wild, unified, framework, jointly, generates, synchronized, facial, speech, enabling, visual, within, model, youngjung, jaeseok, jeong, mingi, kwon, joint, fabrication, aware, turns, natural, language, prompts, into, printable, global, geometry, textures, braille, blind, low, vision, users, visually, impaired, show, all, drs, contribution, advising, publications, oct, jan, visiting, pittsburgh, accessibility, carnegie, mellon, may, remote, while, was, mode, now, part, roblox, yicong, hong, dec, consistent, dynamics, experience, 2019, summa, cum, laude, pohang, science, technology, integrated, adviser, national, education, joining, followed, fall, hosted, about, gave, talks, daydream, pika, both, are, started, internship, excited, share, spent, lot, effort, earlier, year, were, huge, thanks, authors, news, phd, interested, precise, simulation, creation, most, recently, been, how, learn, preserve, state, feel, reach, out, via, email, discussions, collaborations, eye, you, call, alex, 신중혁,

Text of the page (random words):
joonghyuk joonghyuk shin you can also call me alex 신중혁 joonghyuk at snu ac kr eye hi i am a cs phd student at snu i m interested in building fast interactive generative world models for precise simulation and content creation most recently i ve been working with xun huang on how ar video world models can learn and preserve state feel free to reach out via email for discussions or collaborations news more 2026 06 started my internship at nvidia in santa clara also excited to share that two papers i spent a lot of effort on text2tactilegraphics from my cmu visit and jam flow from earlier last year were accepted to eccv 2026 huge thanks to my co authors 2026 01 two papers motionstream and drpose are accepted to iclr 2026 motionstream is accepted as an oral 2025 12 i will be working at a stealth startup from feb 2026 and at nvidia spatial intelligence lab from summer 2026 both on improving video world models 2025 11 motionstream is on arxiv gave talks at pika ai daydream ai and a stealth startup 2025 06 a paper about text based image editing is accepted in iccv 2025 2024 10 i will be joining adobe research in san francisco as an intern on eli shechtman s team in summer 2025 followed by a research visit to cmu s robotics institute in fall 2025 hosted by jun yan zhu education more seoul national university snu sep 2023 present cse integrated m s and ph d adviser jaesik park pohang university of science and technology postech feb 2019 feb 2023 cse b s summa cum laude experience more nvidia spatial intelligence lab june 2026 dec 2026 research scientist intern santa clara ca exploring video world models with consistent stateful dynamics morpheus ai feb 2026 may 2026 research intern remote from seoul worked with xun huang and yicong hong on stateful video world models while morpheus ai was in stealth mode now part of roblox carnegie mellon university robotics institute oct 2025 jan 2026 visiting ph d student jun yan s lab pittsburgh pa worked with ruihan gao ava pun wenzhen yuan and jun yan zhu on generative ai for accessibility text2tactilegraphics adobe research june 2025 sep 2025 research scientist intern san francisco ca worked with xun huang zhengqi li richard zhang jun yan zhu and eli shechtman on building fast and interactive video generative models motionstream publications equal contribution equal advising show all tl drs text based tactile graphics generation for the visually impaired ruihan gao joonghyuk shin ava pun jaesik park wenzhen yuan jun yan zhu eccv 2026 paper project page code models data tl dr we introduce text2tactilegraphics a fabrication aware generative system that turns natural language prompts into 3d printable 2 5d tactile graphics with global geometry tactile textures and braille for blind and low vision users jam flow joint audio motion synthesis with flow matching mingi kwon joonghyuk shin jaeseok jeong jaesik park youngjung uh eccv 2026 paper project page tl dr we present a unified framework that jointly generates synchronized facial motion and speech using flow matching and mm dit enabling diverse audio visual synthesis tasks within a single model direct reward fine tuning on poses for single image to 3d human in the wild seunguk do minwoo huh joonghyuk shin jaesik park iclr 2026 paper project page code tl dr we introduce drpose a direct reward fine tuning method that improves pose accuracy in single view 3d human reconstruction especially for dynamic and acrobatic poses without requiring expensive 3d human assets motionstream real time video generation with interactive motion controls joonghyuk shin zhengqi li richard zhang jun yan zhu jaesik park eli shechtman xun huang iclr 2026 oral 1 13 paper project page post code tl dr motionstream is a streaming causal real time and long duration video generation system with motion controls operating at 30 fps on a single h100 gpu unlocking new possibilities for interactive content generation exploring multimodal diffusion transformers for enhanced prompt based image editing joonghyuk shin alchan hwang yujin kim daneul kim jaesik park iccv 2025 paper project page code tl dr we perform a systematic analysis of mm dit s bidirectional attention mechanism and introduce a robust prompt based editing method working across diverse mm dit models sd3 series and flux instantdrag improving interactivity in drag based image editing joonghyuk shin daehyeon choi jaesik park siggraph asia 2024 paper project page code 230 tl dr we present instantdrag an optimization free pipeline for fast interactive drag based image editing that requires only an image and drag instruction as input learning from real world video datasets fill up balancing long tailed data with generative models joonghyuk shin minguk kang jaesik park arxiv 2023 paper project page tl dr we propose a two stage method for long tailed lt recognition using textual inverted tokens to synthesize images achieving sota results on standard benchmarks when trained from scratch studiogan a taxonomy and benchmark of gans for image synthesis minguk kang joonghyuk shin jaesik park tpami 2023 paper code 3500 tl dr we present studiogan a comprehensive library for gans that reproduces over 30 popular models providing extensive benchmarks and a fair evaluation protocol for image synthesis tasks personal more i am a big fan of baseball i played for postech baseball team tachyons for 5 years as a captain and a catcher i love animals i live with a dog named poby i also like pokemon travelling and fifa video games last updated on june 2026 with face looker