Meta tags:
author= Gengshan Yang;
description= Gengshan Yang - Member of Technical Staff at World Labs. PhD in Robotics from CMU. Research in 4D reconstruction, world models, and 3D vision.;
Headings (most frequently used words):
gengshan, yang, 杨庚山, product, code, research, show, selected, activities, fun,
Text of the page (most frequently used words):
yang (26), #gengshan (24), and (21), deva (18), ramanan (18), from (18), for (18), videos (15), the (14), reconstruction (13), cvpr (11), video (11), flow (8), models (7), 2023 (7), optical (6), neural (6), motion (6), given (6), 2021 (6), monocular (6), zhang (6), casual (6), nbsp (6), depth (5), over (5), time (5), frames (5), articulated (5), shape (5), with (5), view (5), 2024 (5), 2019 (4), single (4), end (4), framework (4), that (4), images (4), neurips (4), scene (4), oral (4), two (4), rigid (4), pixel (4), model (4), synthesis (4), human (4), 2025 (4), 2026 (4), hao (4), world (4), research (4), 2022 (3), multi (3), image (3), matching (3), high (3), enables (3), test (3), using (3), moving (3), learning (3), liu (3), dense (3), deformable (3), object (3), animatable (3), aware (3), generative (3), map (3), jun (3), yan (3), zhu (3), kangle (3), deng (3), dynamic (3), pose (3), views (3), camera (3), tactile (3), diffusion (3), workshop (2), problem (2), whose (2), output (2), such (2), distributions (2), real (2), stereo (2), correspondences (2), several (2), volumetric (2), parameters (2), volume (2), correspondence (2), architecture (2), expansion (2), which (2), 2020 (2), into (2), multiple (2), segment (2), differentiable (2), rendering (2), segmentation (2), priors (2), forrester (2), cole (2), daniel (2), vlasic (2), varun (2), jampani (2), deqing (2), sun (2), ners (2), optimizes (2), shubham (2), tulsiani (2), viser (2), surface (2), capturing (2), banmo (2), reconstructs (2), andrea (2), conditional (2), learns (2), jeff (2), tan (2), fields (2), rac (2), wang (2), total (2), recon (2), rgbd (2), embodied (2), iccv (2), ppr (2), zachary (2), manchester (2), john (2), shuo (2), legged (2), slam (2), sensing (2), generation (2), prior (2), interactive (2), behavior (2), agent (2), pad3r (2), trained (2), flex4dhuman (2), via (2), arxiv (2), jen (2), cheng (2), tracing (2), space (2), visible (2), geometry (2), product (2), google (2), credits, materialize, css, jon, barron, carnegie, mellon, jazz, choir, fall, performance, 4th, 5th, cv4animals, vision, activities, fun, cast, continuous, regression, discrete, binary, classification, occupancy, probabilities, voxel, grid, reliably, efficiently, captures, modal, ambiguous, cases, iros, peiyun, inferring, michael, happold, address, res, imagery, searches, incrementally, coarse, fine, hierarchy, proposed, joshua, manela, hierarchical, deep, resolution, introduce, simple, modifications, layers, significantly, reduces, computation, adaptation, cost, size, converges, much, faster, networks, describe, upgrade, reveals, changes, elements, things, closer, will, get, bigger, upgrading, through, analyze, how, decompose, background, bodies, propose, groups, motions, template, free, approach, combining, data, driven, william, freeman, huiwen, chang, lasr, unposed, same, instance, textured, along, illumination, jason, reflectance, surfaces, sparse, wild, long, short, jointly, shapes, embedding, establish, spotlight, specific, embeddings, hanbyul, joo, vedaldi, natalia, neverova, minh, building, many, controllable, label, edge, our, synthesize, consistent, different, viewpoints, distill, offline, optimized, nerfs, appearance, predictors, efficient, distilling, category, level, disentangles, morphology, allows, retargeting, dinesh, reddy, chaoyang, reconstructing, categories, explains, compositional, including, 3rd, person, bird, eye, extreme, chonghyuk, song, builds, environment, physical, configurations, satisfy, dynamics, contact, constraints, physically, plausible, transfer, robots, icra, swaminathan, gurumurthy, arun, bishop, slomo, general, system, robot, imitation, gaussian, splatting, precise, tracking, fidelity, jonathon, luiten, sebastian, scherer, krishna, murthy, jatavallabhula, jay, karhade, nikhil, keetha, splatam, splat, track, gaussians, rgb, enhance, geometric, details, wenzhen, yuan, ruihan, gao, dreamfusion, exploiting, cloth, interactions, enabled, layer, field, fuses, generic, normal, 3dv, donglai, xiang, dressrecon, freeform, learn, grounded, iclr, angjoo, kanazawa, shunsuke, saito, bajcsy, sim, longitudinal, freely, instances, regressors, siggraph, asia, jia, bin, huang, songwei, yiran, haowen, ting, hsuan, liao, turns, few, syncronized, auto, regressive, without, skeleton, inputs, jenq, neng, hwang, yipeng, flexible, predicts, ray, points, both, occluded, transformer, narendra, ahuja, christoph, lassner, ben, mildenhall, hua, paul, mohamed, banani, aligned, beyond, hover, animation, show, selected, pytorch, code, marble, multimodal, blog, email, github, scholar, completed, phd, robotics, cmu, advised, thesis, also, interned, fair, argo, received, qualcomm, innovation, fellowship, here, worked, labs, early, driving, iterations, intersection, large, was, scientist, meta, 杨庚山,
Text of the page (random words):
gengshan yang gengshan yang 杨庚山 i worked at world labs from 2024 to early 2026 driving research and product iterations at the intersection of large 3d models and generative models prior to that i was a research scientist at meta i completed my phd in robotics at cmu advised by deva ramanan here is my thesis i ve also interned at google research fair and argo ai i received the qualcomm innovation fellowship 2021 google scholar nbsp nbsp github nbsp nbsp x nbsp nbsp email product marble a multimodal world model blog code a framework for 4d reconstruction from videos in pytorch research show selected hover over images for animation 2026 world tracing generative pixel aligned geometry beyond the visible hao zhang mohamed el banani jen hao cheng paul zhang yi hua ben mildenhall christoph lassner narendra ahuja gengshan yang arxiv 2026 world tracing predicts a ray of camera space 3d points capturing both visible and occluded geometry via a diffusion transformer trained with pixel space flow matching flex4dhuman flexible multi view video diffusion for 4d human reconstruction jen hao cheng yipeng wang hao zhang gengshan yang jenq neng hwang arxiv 2026 flex4dhuman turns a single video or a few videos to dense multi view syncronized videos via auto regressive diffusion without skeleton or depth inputs 2025 pad3r pose aware dynamic 3d reconstruction from casual videos ting hsuan liao haowen liu yiran xu songwei ge gengshan yang jia bin huang siggraph asia 2025 pad3r reconstructs freely moving instances with test time trained pose regressors agent to sim learning interactive behavior models from casual longitudinal videos gengshan yang andrea bajcsy shunsuke saito angjoo kanazawa iclr 2025 we learn interactive behavior models of an agent grounded in 3d from casual videos dressrecon freeform 4d human reconstruction from monocular video jeff tan donglai xiang shubham tulsiani deva ramanan gengshan yang 3dv 2025 oral dynamic 3d human with cloth object interactions from a single video enabled by a two layer motion field that fuses 3d human prior and generic pixel priors e g normal flow 2024 tactile dreamfusion exploiting tactile sensing for 3d generation ruihan gao kangle deng gengshan yang wenzhen yuan jun yan zhu neurips 2024 using tactile sensing to enhance geometric details for 3d generation splatam splat track map 3d gaussians for dense rgb d slam nikhil keetha jay karhade krishna murthy jatavallabhula gengshan yang sebastian scherer deva ramanan jonathon luiten cvpr 2024 3d gaussian splatting for slam enables precise camera tracking and high fidelity reconstruction using an rgbd camera 2023 slomo a general system for legged robot motion imitation from casual videos john z zhang shuo yang gengshan yang arun bishop swaminathan gurumurthy deva ramanan zachary manchester ra l 2023 icra 2024 an end to end motion transfer framework from monocular videos to legged robots ppr physically plausible reconstruction from monocular videos gengshan yang shuo yang john z zhang zachary manchester deva ramanan iccv 2023 oral given monocular videos ppr builds 4d models of the object and the environment whose physical configurations satisfy dynamics and contact constraints total recon deformable scene reconstruction for embodied view synthesis chonghyuk song gengshan yang kangle deng jun yan zhu deva ramanan iccv 2023 total recon explains an rgbd video with compositional 4d neural fields which enables extreme view synthesis including embodied views 3rd person views and bird s eye views rac reconstructing animatable categories from videos gengshan yang chaoyang wang n dinesh reddy deva ramanan cvpr 2023 rac learns category level deformable 3d models from monocular videos it disentangles morphology and motion and allows for motion retargeting distilling neural fields for real time articulated shape reconstruction jeff tan gengshan yang deva ramanan cvpr 2023 we distill offline optimized dynamic nerfs into efficient video shape pose and appearance predictors 3d aware conditional image synthesis kangle deng gengshan yang deva ramanan jun yan zhu cvpr 2023 a 3d aware conditional generative model for controllable image synthesis given a 2d label map such as a segmentation or edge map our model learns to synthesize images consistent from different viewpoints 2022 banmo building animatable 3d neural models from many casual videos gengshan yang minh vo natalia neverova deva ramanan andrea vedaldi hanbyul joo cvpr 2022 oral given casual videos capturing a deformable object banmo reconstructs an animatable 3d model in a differentiable volume rendering framework 2021 viser video specific surface embeddings for articulated 3d shape reconstruction gengshan yang deqing sun varun jampani daniel vlasic forrester cole ce liu deva ramanan neurips 2021 spotlight given a long video or multiple short videos viser jointly optimizes articulated 3d shapes and a pixel surface embedding to establish dense correspondences over video frames ners neural reflectance surfaces for sparse view 3d reconstruction in the wild jason y zhang gengshan yang shubham tulsiani deva ramanan neurips 2021 given several 8 16 unposed images of the same instance ners optimizes for a textured 3d reconstruction along with the illumination parameters at test time lasr learning articulated shape reconstruction from a monocular video gengshan yang deqing sun varun jampani daniel vlasic forrester cole huiwen chang deva ramanan william t freeman ce liu cvpr 2021 a template free approach for articulated shape reconstruction from a single video by combining differentiable rendering and data driven correspondence and segmentation priors learning to segment rigid motions from two frames gengshan yang deva ramanan cvpr 2021 we analyze how to decompose two frames into a rigid background and multiple moving rigid bodies and propose a neural architecture to segment rigid motion groups given two frames 2020 upgrading optical flow to 3d scene flow through optical expansion gengshan yang deva ramanan cvpr 2020 oral we describe a neural architecture to upgrade 2d optical flow to 3d scene flow using optical expansion which reveals changes in depth of scene elements over frames e g things moving closer will get bigger 2019 volumetric correspondence networks for optical flow gengshan yang deva ramanan neurips 2019 we introduce several simple modifications to the optical flow volumetric layers that 1 significantly reduces computation and parameters 2 enables test time adaptation of cost volume size and 3 converges much faster hierarchical deep stereo matching on high resolution images gengshan yang joshua manela michael happold deva ramanan cvpr 2019 to address the problem of real time stereo matching on high res imagery an end to end framework that searches for correspondences incrementally over a coarse to fine hierarchy is proposed inferring distributions over depth from a single image gengshan yang peiyun hu deva ramanan iros 2019 we cast the continuous problem of depth regression as discrete binary classification whose output is the occupancy probabilities on a 3d voxel grid such output reliably and efficiently captures multi modal depth distributions in ambiguous cases activities fun the 4d vision workshop at cvpr 25 26 the 4th and 5th cv4animals workshop at cvpr 24 25 carnegie mellon jazz choir fall 2022 performance credits to jon barron materialize css
|
