Meta tags:

Headings (most frequently used words):

3d, in, perception, models, visual, with, world, spatial, reasoning, out, of, images, image, text, and, token, no, for, georgia, gkioxari, self, improving, machine, glab, members, research, highlights, teaching, at, caltech, join, us, 2d, tools, sight, mind, evaluating, state, evolution, video, sam, 3dfy, anything, same, or, not, enhancing, vision, language, conversational, segmentation, grounding, abstract, concepts, scalable, supervision, aligning, structure, by, steer3d, feedforward, editing, via, steerable, to, labels, problem, training, reasoners, multimodal, verifiers, find3d, find, any, part, vadar, agentic, ai, dynamic, api, omni3d, large, benchmark, model, object, detection, the, wild, pytorch3d, mask, cnn,

Text of the page (most frequently used words):
and (20), perception (15), the (13), paper (13), georgia (13), #gkioxari (13), phd (11), project (11), #models (10), code (10), 2026 (9), reasoning (7), student (7), with (6), caltech (6), for (6), cvpr (6), visual (6), spatial (6), world (6), vision (5), learning (5), research (5), 2024 (5), images (5), award (5), this (4), large (4), language (4), damiano (4), marsili (4), 2025 (4), ziqi (4), that (4), 148 (3), spring (3), tools (3), data (3), yisong (3), yue (3), training (3), aadarsh (3), sahoo (3), vansh (3), tibrewal (3), understanding (3), was (3), mention (2), your (2), statement (2), email (2), department (2), prospective (2), students (2), post (2), work (2), read (2), 101 (2), winter (2), 2023 (2), iccv (2), nikhila (2), ravi (2), justin (2), johnson (2), library (2), deep (2), pytorch3d (2), model (2), rohun (2), agrawal (2), highlight (2), hongqiao (2), chen (2), text (2), image (2), structure (2), token (2), demo (2), segmentation (2), grounding (2), concepts (2), sam (2), team (2), out (2), state (2), researcher (2), time (2), develop (2), continually (2), systems (2), more (2), from (2), about (2), 2021 (2), women (2), pami (2), apply, directly, name, purpose, separate, needed, cms, interested, computer, representation, short, docs, you, wish, please, information, undergrads, grads, join, teaching, kaiming, piotr, dollár, ross, girshick, 2017, marr, prize, mask, cnn, jeremy, reizenstein, david, novotny, taylor, gordon, wan, yen, open, source, pytorch, garrick, brazil, abhinav, kumar, julian, straub, omni3d, benchmark, object, detection, wild, vadar, agentic, dynamic, api, find3d, find, any, part, iclr, labels, problem, reasoners, multimodal, verifiers, steer3d, feedforward, editing, via, steerable, aligning, conversational, abstract, scalable, supervision, aditya, mehta, ryan, lin, same, not, enhancing, oral, 3dfy, anything, best, honorable, blog, mengzhan, liufu, eccv, sight, mind, evaluating, evolution, video, highlights, columbia, alumni, undergrad, incoming, ines, kim, doc, neehar, peri, ilona, demler, raphi, kang, glab, members, agents, reason, space, scene, reconstruction, recognition, algorithms, improve, closing, loop, between, allowing, critique, augment, curate, their, own, experience, they, can, acquire, stronger, general, over, believe, path, toward, persistent, unlock, new, capabilities, across, both, reconstructing, semantic, group, builds, interpret, get, better, doing, requires, thiking, optimization, tasks, self, improving, machine, recipient, teammates, received, detectron, suite, named, one, influential, advancing, 2019, nominated, awards, 2020, venturebeat, rework, mark, everingham, amazon, okawa, google, faculty, young, packard, fellow, assistant, professor, previously, scientist, meta, completed, undergraduate, studies, greece, where, worked, petros, maragos, ntua, jitendra, malik, berkeley, fair, computing, mathematical, sciences,

Text of the page (random words):
georgia gkioxari georgia gkioxari i am an assistant professor in the computing mathematical sciences department at caltech previously i was a research scientist at meta s fair team i completed my phd at uc berkeley with jitendra malik and my undergraduate studies at ntua greece where i worked with petros maragos i am a packard fellow 2025 the recipient of the pami young researcher award 2021 a google faculty award 2024 the okawa research award 2024 and the amazon research award 2024 my teammates and i received the pami mark everingham award 2021 for the detectron library suite i was named one of 30 influential women advancing ai in 2019 by rework and was nominated for the women in ai awards in 2020 by venturebeat read more about me and my work in this q a self improving machine perception my group builds visual systems that interpret the world from images and continually get better at doing so this requires re thiking optimization training data and tasks we develop visual perception models that unlock new capabilities across both 2d perception and 3d perception from understanding images and reconstructing 3d structure to grounding and reasoning about semantic concepts we develop algorithms that continually improve spatial reasoning by closing the loop between models and data allowing systems to critique augment and re curate their own training experience so they can acquire stronger and more general visual understanding over time we believe this is the path toward state persistent world models 2d perception recognition and segmentation in images 3d perception 3d scene reconstruction and understanding world models spatial reasoning world model agents that reason in space and time tools tools for 3d deep learning pytorch3d glab members ziqi ma phd student damiano marsili phd student aadarsh sahoo phd student raphi kang phd student ilona demler phd student vansh tibrewal phd student neehar peri post doc ines kim incoming phd student hongqiao chen undergrad researcher alumni rohun agrawal bs caltech 2025 phd columbia vansh tibrewal bs caltech 2026 phd caltech research highlights world models out of sight out of mind evaluating state evolution in video world models ziqi ma mengzhan liufu georgia gkioxari eccv 2026 paper project blog 3d perception best paper honorable mention sam 3d 3dfy anything in images sam 3d team cvpr 2026 oral paper project code demo spatial reasoning same or not enhancing visual perception in vision language models damiano marsili aditya mehta ryan y lin georgia gkioxari cvpr 2026 highlight paper project code 2d perception conversational image segmentation grounding abstract concepts with scalable supervision aadarsh sahoo georgia gkioxari cvpr 2026 paper project demo code 3d perception aligning text images and 3d structure token by token aadarsh sahoo vansh tibrewal georgia gkioxari cvpr 2026 paper project code 3d perception steer3d feedforward 3d editing via text steerable image to 3d ziqi ma hongqiao chen yisong yue georgia gkioxari paper project code spatial reasoning no labels no problem training visual reasoners with multimodal verifiers damiano marsili georgia gkioxari iclr 2026 paper project code 3d perception find3d find any part in 3d ziqi ma yisong yue georgia gkioxari iccv 2025 highlight paper project code spatial reasoning vadar visual agentic ai for spatial reasoning with a dynamic api damiano marsili rohun agrawal yisong yue georgia gkioxari cvpr 2025 paper project code data 3d perception omni3d a large benchmark and model for 3d object detection in the wild garrick brazil abhinav kumar julian straub nikhila ravi justin johnson georgia gkioxari cvpr 2023 paper project code tools pytorch3d nikhila ravi jeremy reizenstein david novotny taylor gordon wan yen lo justin johnson georgia gkioxari open source library for 3d deep learning in pytorch paper project code 2d perception mask r cnn kaiming he georgia gkioxari piotr dollár ross girshick iccv 2017 marr prize paper teaching at caltech ee cs 148 spring 2023 large language vision models ee cs 148 spring 2024 large language vision models ee cs 148 spring 2026 large language vision models cs 101 winter 2024 learning 3d cs 101 winter 2026 learning 3d join us caltech students undergrads grads if you wish to work with me please read this information prospective post docs interested in computer vision 3d representation learning or perception email me your cv and a short research statement prospective phd students apply directly to the cms department and mention my name in your statement of purpose no separate email needed