Meta tags:
description=TextureDreamer: Image-guided Texture Synthesis through Geometry-aware Diffusion;
keywords=Texture Transfer, Texture Synthesis;

Headings (most frequently used words):

nbsp, texture, of, comparisons, with, baselines, texturedreamer, geometry, aware, results, image, guided, synthesis, through, diffusion, cvpr, 2024, yu, ying, yeh1, jia, bin, huang2, changil, kim3, lei, xiao3, thu, nguyen, phuoc3, numair, khan3, cheng, zhang3, manmohan, chandraker1, carl, marshall3, zhao, dong3, zhengqin, li3, 1university, california, san, diego, 2university, maryland, college, park, 3meta, arxiv, transfers, photorealistic, high, fidelity, and, textures, from, images, to, arbitrary, 3d, meshes, abstract, method, bibtex, sofa, plush, mug, bed, cross, category, transfer, relighting, ablation, study, diversity,

Text of the page (most frequently used words):
texture (25), and (21), with (21), #images (20), input (15), nbsp (13), ours (10), mesh (10), bed (8), latentpaint (8), from (7), #texturedreamer (7), geometry (7), our (7), method (7), the (7), style (7), aware (6), synthesis (5), personalized (5), image (4), guided (4), through (4), diffusion (4), can (4), armchair (4), bear (4), sofa (4), mug (4), transfer (4), compare (4), latent (4), paint (4), comparisons (4), baselines (4), arbitrary (4), textures (4), arxiv (3), cfg (3), controlnet (3), light (3), shapes (3), chair (3), pillow (3), category (3), black (3), score (3), distillation (3), for (3), yeh (2), ying (2), huang (2), jia (2), bin (2), kim (2), changil (2), xiao (2), lei (2), nguyen (2), phuoc (2), thu (2), khan (2), numair (2), zhang (2), cheng (2), chandraker (2), manmohan (2), marshall (2), carl (2), dong (2), zhao (2), 2024 (2), diversity (2), model (2), sds (2), ablation (2), relighting (2), cup (2), plate (2), pikachu (2), double (2), three (2), seat (2), meshes (2), red (2), plaid (2), pink (2), cross (2), plush (2), results (2), brdf (2), field (2), optimized (2), pgsd (2), high (2), detailed (2), quality (2), real (2), categories (2), highly (2), objects (2), creation (2), methods (2), university (2), this, website, template, partially, borrowed, nerfies, article, yeh2024texturedreamer, title, author, others, journal, preprint, 2401, 09416, year, bibtex, synthesize, diverse, patterns, same, set, camera, encoder, updated, lora, removed, 100, depth, green, gray, back, study, single, white, colorful, brown, head, cloak, king, size, strips, floral, yellow, teapot, bowl, orangutan, tokage, cat, flowers, given, first, obtain, finetuning, spatially, varying, bidirectional, reflectance, distribution, then, geometric, after, optimization, finished, resolution, maps, corresponding, albedo, metallic, roughness, extracted, dreambooth, core, idea, draws, inspiration, recent, advancements, diffuse, models, including, modeling, information, extraction, variational, appearance, explicit, guidance, integration, several, essential, modifications, substantially, improve, experiments, spanning, different, show, that, successfully, realistic, semantic, meaningful, surpassing, visual, previous, state, art, pivotal, challenge, vision, graphics, industrial, companies, hire, experienced, artists, manually, craft, assets, classical, require, densely, sampled, views, accurately, aligned, while, learning, based, are, confined, specific, within, dataset, contrast, intricate, world, environments, only, few, casually, captured, potentially, significantly, democratizing, present, novel, relightable, small, number, target, across, abstract, transfers, photorealistic, fidelity, cvpr, meta, maryland, college, park, california, san, diego, zhengqin,

Text of the page (random words):
texturedreamer image guided texture synthesis through geometry aware diffusion texturedreamer image guided texture synthesis through geometry aware diffusion cvpr 2024 yu ying yeh 1 3 nbsp jia bin huang 2 3 nbsp changil kim 3 nbsp lei xiao 3 nbsp thu nguyen phuoc 3 nbsp numair khan 3 nbsp cheng zhang 3 nbsp manmohan chandraker 1 nbsp carl s marshall 3 nbsp zhao dong 3 nbsp zhengqin li 3 nbsp 1 university of california san diego nbsp 2 university of maryland college park nbsp 3 meta arxiv results texturedreamer transfers photorealistic high fidelity and geometry aware textures from 3 5 images to arbitrary 3d meshes abstract we present texturedreamer a novel image guided texture synthesis method to transfer relightable textures from a small number of input images 3 to 5 to target 3d shapes across arbitrary categories texture creation is a pivotal challenge in vision and graphics industrial companies hire experienced artists to manually craft textures for 3d assets classical methods require densely sampled views and accurately aligned geometry while learning based methods are confined to category specific shapes within the dataset in contrast texturedreamer can transfer highly detailed intricate textures from real world environments to arbitrary objects with only a few casually captured images potentially significantly democratizing texture creation our core idea personalized geometry aware score distillation pgsd draws inspiration from recent advancements in diffuse models including personalized modeling for texture information extraction variational score distillation for detailed appearance synthesis and explicit geometry guidance with controlnet our integration and several essential modifications substantially improve the texture quality experiments on real images spanning different categories show that texturedreamer can successfully transfer highly realistic semantic meaningful texture to arbitrary objects surpassing the visual quality of previous state of the art method given 3 5 images we first obtain personalized diffusion model with dreambooth finetuning the spatially varying bidirectional reflectance distribution brdf field is then optimized through personalized geometric aware score distillation pgsd after optimization finished high resolution texture maps corresponding to albedo metallic and roughness can be extracted from the optimized brdf field results sofa plush mug bed cross category relighting ablation diversity comparisons with baselines sofa we compare our method with latent paint and texture input style black chair with flowers red plaid chair with a pillow input images images mesh armchair ours texture latentpaint mesh three seat sofa ours texture latentpaint comparisons with baselines plush we compare our method with latent paint and texture input style cat tokage input images images mesh pikachu ours texture latentpaint mesh orangutan ours texture latentpaint comparisons with baselines mug we compare our method with latent paint and texture input style bowl bear mug input images images mesh cup with plate ours texture latentpaint mesh teapot ours texture latentpaint comparisons with baselines bed we compare our method with latent paint and texture input style black floral bed with a yellow pillow bed with pink and black strips input images images mesh double bed ours texture latentpaint mesh king size bed ours texture latentpaint cross category texture transfer input style brown bear with a pink head cloak red plaid chair with a pillow white mug with colorful shapes input images images meshes armchair three seat sofa double bed single bed pikachu cup with plate relighting of texture input style bear input images images mesh bear light 0 light 1 light 2 ablation study input style green armchair with gray back input images images mesh armchair w o controlnet w controlnet depth sds w o cfg sds cfg 100 w o lora removed personalized model as φ ours cfg 7 5 w o camera encoder ρ updated ours diversity of texture our method can synthesize diverse patterns from the same set of images bibtex article yeh2024texturedreamer title texturedreamer image guided texture synthesis through geometry aware diffusion author yeh yu ying and huang jia bin and kim changil and xiao lei and nguyen phuoc thu and khan numair and zhang cheng and chandraker manmohan and marshall carl s and dong zhao and others journal arxiv preprint arxiv 2401 09416 year 2024 this website template is partially borrowed from nerfies