Meta tags:

Headings (most frequently used words):

dynamic, rodynrf, radiance, camera, on, the, as, and, fields, video, of, poses, estimation, dataset, view, synthesis, results, to, robust, cvpr, 2023, abstract, evaluation, mpi, sintel, novel, nvidia, scene, consistent, geometry, bibtex, takes, casually, captured, input, reconstructs, trajectory, conventional, sfm, system, such, colmap, fails, recover, even, when, using, ground, truth, motion, masks, result, existing, field, methods, that, require, accurate, pose, are, inapplicable, these, challenging, scenes, tackles, this, robustness, problem, showcases, high, fidelity, wide, variety, videos,

Text of the page (most frequently used words):
and (19), the (15), dynamic (14), #radiance (8), methods (8), poses (8), our (7), #camera (7), fields (6), rodynrf (6), method (5), results (5), with (4), existing (4), robustness (4), sfm (4), this (3), robust (3), performs (3), colmap (3), view (3), synthesis (3), estimation (3), that (3), algorithms (3), video (3), motion (3), from (2), liu (2), lun (2), gao (2), chen (2), meuleman (2), andreas (2), tseng (2), hung (2), saraf (2), ayush (2), kim (2), changil (2), chuang (2), yung (2), kopf (2), johannes (2), huang (2), jia (2), bin (2), 2023 (2), reconstructs (2), consistent (2), geometry (2), favorably (2), against (2), state (2), art (2), even (2), can (2), using (2), scene (2), dataset (2), based (2), pose (2), evaluation (2), issue (2), jointly (2), estimating (2), static (2), along (2), parameters (2), focal (2), length (2), field (2), structure (2), accurate (2), these (2), are (2), challenging (2), videos (2), university (2), website, borrowed, nerfies, inproceedings, liu2023robust, author, title, booktitle, proceedings, ieee, cvf, conference, computer, vision, pattern, recognition, year, bibtex, furthermore, without, still, achieve, comparable, ones, novel, nvidia, significantly, better, than, nerf, note, also, learning, visual, odometry, mpi, sintel, addresses, reconstruction, aim, model, time, varying, appearance, however, assume, reliably, estimated, thus, unreliable, often, fail, produce, erroneous, highly, objects, poorly, textured, surfaces, rotating, address, demonstrate, approach, via, extensive, quantitative, qualitative, experiments, show, favorable, performance, over, abstract, takes, casually, captured, input, trajectory, conventional, system, such, fails, recover, when, ground, truth, masks, result, require, inapplicable, scenes, tackles, problem, showcases, high, fidelity, wide, variety, code, supp, arxiv, cvpr, maryland, college, park, kaist, national, taiwan, meta,

Text of the page (random words):
rodynrf robust dynamic radiance fields rodynrf robust dynamic radiance fields yu lun liu chen gao andreas meuleman hung yu tseng ayush saraf changil kim yung yu chuang johannes kopf jia bin huang meta national taiwan university kaist university of maryland college park cvpr 2023 arxiv video supp code evaluation results rodynrf takes a casually captured video as input and reconstructs the camera trajectory and dynamic radiance fields conventional sfm system such as colmap fails to recover camera poses even when using ground truth motion masks as a result existing dynamic radiance field methods that require accurate pose estimation are inapplicable to these challenging dynamic scenes rodynrf tackles this robustness problem and showcases high fidelity dynamic view synthesis results on a wide variety of videos abstract dynamic radiance field reconstruction methods aim to model the time varying structure and appearance of a dynamic scene existing methods however assume that accurate camera poses can be reliably estimated by structure from motion sfm algorithms these methods thus are unreliable as sfm algorithms often fail or produce erroneous poses on challenging videos with highly dynamic objects poorly textured surfaces and rotating camera motion we address this robustness issue by jointly estimating the static and dynamic radiance fields along with the camera parameters poses and focal length we demonstrate the robustness of our approach via extensive quantitative and qualitative experiments our results show favorable performance over the state of the art dynamic view synthesis methods video rodynrf rodynrf addresses the robustness issue of sfm algorithms by jointly estimating the static and dynamic radiance fields along with the camera parameters poses and focal length evaluation of camera poses estimation on the mpi sintel dataset our method performs significantly better than existing nerf based pose estimation methods note that our method also performs favorably against existing learning based visual odometry methods novel view synthesis results on the nvidia dynamic scene dataset our method performs favorably against state of the art methods furthermore even without colmap poses our method can still achieve results comparable with the ones using colmap poses consistent geometry our method reconstructs consistent geometry bibtex inproceedings liu2023robust author liu yu lun and gao chen and meuleman andreas and tseng hung yu and saraf ayush and kim changil and chuang yung yu and kopf johannes and huang jia bin title robust dynamic radiance fields booktitle proceedings of the ieee cvf conference on computer vision and pattern recognition year 2023 this website is borrowed from nerfies