Meta tags:

Headings (most frequently used words):

3d, articulated, articulate, anymesh, objects, mesh, generated, an, object, by, starting, from, real, to, please, select, example, input, open, vocabulary, modeling, abstract, qualitative, results, applications, annotated, handcrafted, policy, learning, sim, ouput, output,

Text of the page (most frequently used words):
#articulated (16), and (11), objects (11), object (8), the (8), #articulate (8), anymesh (8), generated (5), mesh (5), real (4), manipulation (4), from (4), bucket (3), laptop (3), for (3), open (3), vocabulary (3), that (3), our (3), can (3), wang (3), simulation (2), demos (2), policy (2), learning (2), training (2), set (2), experiments (2), categories (2), input (2), please (2), select (2), example (2), starting (2), handcrafted (2), results (2), modeling (2), both (2), parts (2), joints (2), existing (2), which (2), wide (2), range (2), these (2), framework (2), into (2), show (2), including (2), vehicles (2), mit (2), powered, bootstrap, jquery, world, sim, follow, evaluate, two, tasks, augment, include, additional, same, lifting, opening, lid, dexart, applications, name, output, ouput, annotated, qualitative, has, long, been, challenging, problem, since, requires, capture, accurate, surface, geometries, semantically, meaningful, spatially, precise, structures, methods, heavily, depend, data, limited, cabinets, drawers, restricts, their, ability, model, context, address, limitations, propose, automated, able, convert, any, rigid, its, counterpart, manner, given, utilizes, advanced, vision, language, models, visual, prompting, techniques, extract, semantic, information, allowing, segmentation, construction, functional, generate, large, scale, high, quality, tools, toys, mechanical, devices, significantly, expanding, coverage, datasets, additionally, assets, facilitate, acquisition, new, skills, then, transferred, robotic, system, abstract, converts, meshes, retrieved, pipeline, capable, processing, daily, necessities, furniture, even, fictional, objaverse, try, clicking, teaser, see, code, arxiv, umass, amherst, sjtu, cmu, csail, ibm, lab, chuang, gan, zhou, xian, tsun, hsuan, yufei, zhehuan, chen, yian, jincheng, yang, xiaowen, qiu,

Text of the page (random words):
articulate anymesh articulate anymesh open vocabulary 3d articulated objects modeling xiaowen qiu 1 jincheng yang 2 yian wang 1 zhehuan chen 1 yufei wang 3 tsun hsuan wang 4 zhou xian 3 chuang gan 1 5 1 umass amherst 2 sjtu 3 cmu 4 mit csail 5 mit ibm lab arxiv code try clicking objects in the teaser to see articulated results articulate anymesh converts 3d meshes e g retrieved from objaverse into articulated objects our pipeline is capable of processing a wide range of objects including daily necessities furniture vehicles and even fictional objects abstract 3d articulated objects modeling has long been a challenging problem since it requires to capture both accurate surface geometries and semantically meaningful and spatially precise structures parts and joints existing methods heavily depend on training data from a limited set of handcrafted articulated object categories e g cabinets and drawers which restricts their ability to model a wide range of articulated objects in an open vocabulary context to address these limitations we propose articulate anymesh an automated framework that is able to convert any rigid 3d mesh into its articulated counterpart in an open vocabulary manner given a 3d mesh our framework utilizes advanced vision language models and visual prompting techniques to extract semantic information allowing for both the segmentation of object parts and the construction of functional joints our experiments show that articulate anymesh can generate large scale high quality 3d articulated objects including tools toys mechanical devices and vehicles significantly expanding the coverage of existing 3d articulated object datasets additionally we show that these generated assets can facilitate the acquisition of new articulated object manipulation skills in simulation which can then be transferred to a real robotic system qualitative results articulated objects annotated by articulate anymesh starting from a handcrafted 3d mesh please select an example input 3d object ouput articulated mesh articulated objects generated by articulate anymesh starting from an generated 3d object please select an example input generated 3d object output articulated mesh name applications policy learning we follow dexart for articulated object manipulation policy learning and evaluate two tasks opening a laptop lid and lifting a bucket we augment the training set for the bucket and laptop experiments we include additional articulated objects of the same categories generated by articulate anymesh laptop manipulation demos bucket manipulation demos real to sim to real simulation real world powered by bootstrap jquery