Meta tags:
description=Roman Bachmann is a Machine Learning Research Scientist at Apple working on any-to-any multimodal foundation models, tokenization, and world models.;
author=Roman Bachmann;

Headings (most frequently used words):

vision, for, any, of, and, visual, to, learning, multi, from, motion, roman, bachmann, scalable, training, modeling, modalities, 1d, flexible, length, multimodal, models, computer, tasks, 4m, masked, task, 3d, using, capture, pan, tilt, cameras, ml, research, scientist, 2026, 2025, 2024, 2023, 2022, 2021, 2020, 2019, weblica, reproducible, environments, web, agents, modus, decoder, only, diverse, ordered, tokens, enable, efficient, test, time, search, videoflextok, coarse, fine, video, tokenization, how, well, does, gpt, 4o, understand, evaluating, foundation, on, standard, flextok, resampling, images, into, token, sequences, 21, an, model, tens, viper, personalization, generative, via, individual, preference, massively, modality, invariant, odometry, embodied, multimae, modal, autoencoders, clipasso, semantically, aware, object, sketching, omnidata, pipeline, making, mid, level, datasets, scans, binary, neural, networks, the, bayesian, rule, with, unknown, orientation, global, estimation, automatic, in, alpine, skiing, deep,

Text of the page (most frequently used words):
roman (21), bachmann (21), arxiv (15), amir (14), and (13), zamir (13), code (12), website (12), vision (10), fatih (8), kar (8), #computer (7), for (7), david (7), afshin (7), dehghan (7), any (6), oğuzhan (6), 2026 (6), learning (5), conference (5), mizrahi (5), models (5), research (5), dataset (4), visual (4), gao (4), multimodal (4), icml (4), jesse (4), allardice (4), scientist (4), pdf (3), the (3), science (3), helge (3), rhodin (3), pascal (3), fua (3), motion (3), best (3), 2019 (3), from (3), international (3), scalable (3), multi (3), 2022 (3), andrei (3), atanov (3), 2023 (3), 2024 (3), ali (3), foundation (3), world (3), award (2), skiing (2), non (2), peer (2), reviewed (2), jörg (2), spörri (2), capture (2), using (2), presentation (2), paper (2), european (2), global (2), pan (2), tilt (2), cameras (2), 2020 (2), machine (2), training (2), 2021 (2), task (2), masked (2), spotlight (2), neurips (2), oguzhan (2), teresa (2), yeo (2), mingfei (2), modeling (2), garjani (2), griffiths (2), tasks (2), modalities (2), 2025 (2), flexible (2), length (2), tokenization (2), zhitong (2), mingqiao (2), enable (2), epfl (2), apple (2), reasoning (2), won, first, place, young, investigator, competition, 8th, int, congress, automatic, alpine, deep, voted, third, central, seminar, graphics, estimation, selected, oral, with, unknown, orientation, xiangming, meng, mohammad, emtiyaz, khan, binary, neural, networks, bayesian, rule, ainaz, eftekhar, alexander, sax, jitendra, malik, omnidata, pipeline, making, mid, level, datasets, scans, winner, siggraph, yael, vinker, ehsan, pajouheshgar, jessica, amit, haim, bermano, daniel, cohen, ariel, shamir, clipasso, semantically, aware, object, sketching, multimae, modal, autoencoders, pattern, recognition, marius, memmel, modality, invariant, odometry, embodied, massively, eccv, sogand, salehi, mahdi, shafiei, viper, personalization, generative, via, individual, preference, jiaming, model, tens, enrico, fini, elmira, amirloo, alaaeldin, nouby, flextok, resampling, images, into, token, sequences, iclr, rahul, ramachandran, how, well, does, gpt, understand, evaluating, standard, devon, hjelm, peter, videoflextok, coarse, fine, video, parham, rezaei, nataša, jovanović, ordered, tokens, efficient, test, time, search, zhaochong, xian, liu, françois, fleuret, chuan, zadeh, serge, belongie, modus, decoder, only, diverse, yuanzheng, gong, anders, boesen, lindbo, larsen, weblica, reproducible, environments, web, agents, previously, was, phd, student, advised, received, degree, data, where, also, completed, during, studies, interned, riken, aip, vilab, focused, building, modelling, goal, build, adaptable, priors, that, quick, understanding, environment, allow, out, sight, github, google, scholar, email, generation,

Text of the page (random words):
roman bachmann roman bachmann ml research scientist multimodal foundation models tokenization generation world models email google scholar github roman bachmann ml research scientist hi i m roman bachmann a machine learning research scientist at apple my research is focused on building scalable any to any multimodal foundation models for world modelling and visual reasoning my goal is to build adaptable world priors that enable quick understanding of the environment and allow for global out of sight reasoning previously i was an epfl phd student at vilab advised by amir zamir i received my m sc degree in data science at epfl where i also completed my b sc in computer science during my studies i interned as a research scientist at apple and riken aip 2026 weblica scalable and reproducible training environments for visual web agents oğuzhan fatih kar roman bachmann yuanzheng gong anders boesen lindbo larsen afshin dehghan arxiv 2026 arxiv modus decoder only any to any modeling of diverse modalities mingqiao ye zhaochong an zhitong gao xian liu françois fleuret chuan li amir zadeh serge belongie afshin dehghan jesse allardice david mizrahi oğuzhan fatih kar roman bachmann amir zamir icml 2026 pdf website code dataset 1d ordered tokens enable efficient test time search zhitong gao parham rezaei ali cy mingqiao ye nataša jovanović jesse allardice afshin dehghan amir zamir roman bachmann oğuzhan fatih kar icml 2026 arxiv website code videoflextok flexible length coarse to fine video tokenization andrei atanov jesse allardice roman bachmann oğuzhan fatih kar r devon hjelm david griffiths peter fu afshin dehghan amir zamir icml 2026 spotlight arxiv website code 2025 how well does gpt 4o understand vision evaluating multimodal foundation models on standard computer vision tasks rahul ramachandran ali garjani roman bachmann andrei atanov oğuzhan fatih kar amir zamir iclr 2026 arxiv website code flextok resampling images into 1d token sequences of flexible length roman bachmann jesse allardice david mizrahi enrico fini oğuzhan fatih kar elmira amirloo alaaeldin el nouby amir zamir afshin dehghan icml 2025 arxiv website code 2024 4m 21 an any to any vision model for tens of tasks and modalities roman bachmann oguzhan fatih kar david mizrahi ali garjani mingfei gao david griffiths jiaming hu afshin dehghan amir zamir neurips 2024 arxiv website code viper visual personalization of generative models via individual preference learning sogand salehi mahdi shafiei teresa yeo roman bachmann amir zamir eccv 2024 arxiv website code 2023 4m massively multimodal masked modeling david mizrahi roman bachmann oguzhan fatih kar teresa yeo mingfei gao afshin dehghan amir zamir neurips 2023 spotlight arxiv website code modality invariant visual odometry for embodied vision marius memmel roman bachmann amir zamir 2023 conference on computer vision and pattern recognition arxiv website code 2022 multimae multi modal multi task masked autoencoders roman bachmann david mizrahi andrei atanov amir zamir 2022 european conference on computer vision arxiv website code clipasso semantically aware object sketching yael vinker ehsan pajouheshgar jessica y bo roman bachmann amit haim bermano daniel cohen or amir zamir ariel shamir siggraph 2022 best paper award winner arxiv website code 2021 omnidata a scalable pipeline for making multi task mid level vision datasets from 3d scans ainaz eftekhar alexander sax roman bachmann jitendra malik amir zamir 2021 international conference on computer vision arxiv website 2020 training binary neural networks using the bayesian learning rule xiangming meng roman bachmann and mohammad emtiyaz khan 2020 international conference on machine learning arxiv code 2019 motion capture from pan tilt cameras with unknown orientation roman bachmann jörg spörri pascal fua and helge rhodin 2019 international conference on 3d vision selected for oral presentation arxiv dataset global motion estimation from pan tilt cameras roman bachmann helge rhodin and pascal fua 2019 central european seminar on computer graphics non peer reviewed voted best presentation and third best paper pdf dataset automatic 3d motion capture in alpine skiing using deep learning and computer vision roman bachmann helge rhodin jörg spörri and pascal fua 8th int congress on science and skiing non peer reviewed won first place in the young investigator award competition pdf dataset