Meta tags:
description=Aditya Agarwal s personal academic website;
author=Aditya Agarwal;

Headings (most frequently used words):

for, and, 3d, in, planning, learning, to, video, lipreading, open, world, scene, completion, manipulation, perception, nerf, improved, using, towards, an, publications, news, announcements, academic, services, professional, achievements, scenecomplete, cluttered, real, environments, robot, vector, quantized, feature, fields, fast, semantic, lifting, look, around, enhancing, teleoperation, with, human, like, actuated, neck, conceptgraphs, vocabulary, graphs, edmp, ensemble, of, costs, guided, diffusion, motion, hyp, priors, hypernetwork, disentangling, control, non, prehensile, tabletop, scarp, shape, arbitrary, poses, grasping, faceoff, face, swapping, system, moocs, synthetic, talking, heads, train, humans, at, scale, inr, continuous, representation, space, based, generative, tasks, approaches, challenges, robotic, table, top, rearrangement, personalized, one, shot, als, patient, reed, approach, quickly, bootstrapping, multilingual, acoustic, models,

Text of the page (most frequently used words):
and (70), the (64), for (38), video (22), paper (19), #learning (18), planning (16), aditya (16), agarwal (16), 2023 (14), bipasha (13), sen (13), lipreading (12), from (11), with (11), icra (10), face (10), was (9), research (9), models (9), scene (9), code (9), propose (9), that (9), project (8), training (8), tasks (8), manipulation (8), page (8), perception (8), accepted (7), using (7), swapping (7), open (7), krishna (7), world (7), work (6), two (6), part (6), tabletop (6), jan (6), university (6), based (6), shape (6), completion (6), method (6), madhava (6), india (5), microsoft (5), mar (5), awarded (5), award (5), one (5), aug (5), will (5), 2022 (5), 2024 (5), computer (5), robotics (5), generation (5), representation (5), professors (5), representations (5), nerf (5), our (5), real (5), jawahar (5), vinay (5), namboodiri (5), pose (5), task (5), motion (5), scenes (5), singh (5), robot (5), cvit (4), summer (4), iiit (4), reviewer (4), 2025 (4), performance (4), student (4), shot (4), grant (4), this (4), intern (4), robotic (4), challenge (4), existing (4), datasets (4), their (4), hypernetwork (4), feature (4), semantic (4), lifting (4), embodied (4), advised (4), adaptation (3), building (3), education (3), institute (3), school (3), across (3), rearrangement (3), talk (3), academic (3), through (3), sep (3), completed (3), science (3), nov (3), multilingual (3), acoustic (3), low (3), hyderabad (3), setting (3), may (3), news (3), wacv (3), talking (3), generative (3), arbitrary (3), poses (3), mit (3), travel (3), phd (3), group (3), tomás (3), lozano (3), pérez (3), leslie (3), kaelbling (3), systems (3), liam (3), paull (3), neurips (3), hyp (3), improves (3), approach (3), synthetic (3), data (3), rudrabha (3), mukhopadhyay (3), object (3), brojeshwar (3), bhowmick (3), vishal (3), reddy (3), mandadi (3), over (3), inr (3), poster (3), novel (3), expressions (3), without (3), system (3), multi (3), srinath (3), sridhar (3), gaurav (3), improve (3), valid (3), trajectories (3), teleoperation (3), actuated (3), neck (3), human (3), complex (3), enabling (3), efficiency (3), environments (3), robots (3), only (2), you (2), feb (2), mauritius (2), covered (2), scale (2), its (2), several (2), outstanding (2), selected (2), 000 (2), program (2), all (2), vmware (2), coordinator (2), conducted (2), tutorial (2), sessions (2), multimodal (2), long (2), iit (2), challenges (2), workshops (2), cvpr (2), mitacs (2), calgary (2), pes (2), model (2), resource (2), languages (2), slt (2), joining (2), oct (2), domain (2), bmvc (2), were (2), competition (2), news2 (2), news1 (2), papers (2), tmlr (2), space (2), bangalore (2), week (2), works (2), presented (2), moocs (2), v2v (2), implicit (2), columbia (2), brown (2), ieee (2), attend (2), conference (2), apr (2), spoken (2), against (2), csail (2), pack (2), intelligent (2), lis (2), florian (2), shkurti (2), mila (2), intelligence (2), eecs (2), new (2), tackled (2), proposed (2), mechanism (2), inference (2), slides (2), 2021 (2), towards (2), medical (2), people (2), trained (2), large (2), enable (2), end (2), pipeline (2), incorporating (2), objects (2), configuration (2), target (2), grasp (2), table (2), shapes (2), inrs (2), train (2), prior (2), techniques (2), remarkable (2), rely (2), movements (2), content (2), vision (2), built (2), such (2), any (2), supplementary (2), body (2), methods (2), identity (2), source (2), faceoff (2), features (2), scarp (2), improved (2), control (2), dynamics (2), level (2), arun (2), kallol (2), saha (2), priors (2), hypernetworks (2), resulting (2), quality (2), finetune (2), multiview (2), edmp (2), diverse (2), diffusion (2), specific (2), costs (2), conceptgraphs (2), vocabulary (2), leveraging (2), foundation (2), view (2), demonstrate (2), enhancing (2), vector (2), quantized (2), fields (2), efficient (2), enables (2), pixel (2), aligned (2), masks (2), fast (2), scenecomplete (2), cluttered (2), reason (2), about (2), show (2), publications (2), agents (2), dynamically (2), achieve (2), ended (2), study (2), aware (2), compositional (2), reasoning (2), policy (2), also (2), demonstrations (2), engineering (2), lab (2), where (2), worked (2), prof (2), change, constant, entity, don, play, victim, circumstances, created, forked, modified, theme, pixyll, viraj, prabhu, flagship, deployed, inaugurated, msr, minister, tertiary, press, massively, empowered, classroom, helped, platform, first, 100k, users, media, outlets, delight, your, customer, community, individuals, out, employees, undergo, shoot, company, campus, hiring, available, youtube, winners, among, participating, teams, global, relay, opensource, borathon, professional, achievements, self, supervised, 5th, taking, month, machine, faculties, universities, csedu, jointly, gave, 6th, iclr, iros, siggraph, services, interning, 2016, fully, funded, globalink, bachelor, degree, received, distinction, exceptional, winning, 3rd, place, 1000, international, round, acceptance, rate, journal, attending, google, reserach, featured, weekly, publication, rnd, showcase, synergistic, parameterization, invited, slide, deck, organized, graduate, visit, days, here, generous, 2250, ras, committee, premier, london, jun, full, abstract, presentation, tokyo, japan, icmpc17, apscom7, uncovering, biases, indian, artists, fall, starting, montreal, quebec, artificial, joined, massachusetts, technology, see, orleans, louisiana, atlanta, georgia, corl2023, langrob, tgr, prl, scholar, 1700, announcements, problem, bootstrap, validate, compatibility, multiple, cnns, operating, directly, raw, speech, signals, times, respectively, comparable, wers, rnn, baseline, mlads, anil, kumar, vuppala, mirishkar, sai, ganesh, reed, quickly, bootstrapping, patients, there, primary, issues, had, suffering, disabilities, lacked, words, devised, variational, encoder, technique, adapt, amounts, examples, personalized, als, patient, ros, stacks, manipulate, initial, desired, finger, manipulator, involves, following, steps, registration, estimation, cloud, organization, shankara, narayanan, approaches, top, inspired, recent, parameterizing, neural, encode, videos, learn, these, functions, progressive, clip, regularization, stabilize, shows, many, benchmark, openreview, continuous, hard, hearing, mouth, speaker, understand, developed, upon, tts, generate, language, heads, humans, direction, tackles, pressing, moviemaking, industry, actor, double, preserve, swap, facial, along, background, completing, partial, disentangled, rotationally, equivariant, geometric, tasking, objective, proposals, baselines, short, grasping, framework, disentangles, unknown, pushing, striking, tactile, feedback, explicitly, modeling, consists, components, planner, path, controller, case, dipanjan, das, nomaan, qureshi, dipanwita, guhathakurta, disentangling, non, prehensile, latent, conditioning, generalizable, category, use, estimate, both, weights, resolution, hash, encodings, significant, gains, further, incorporate, denoise, strategy, denoises, images, rendered, nerfs, estimated, finetunes, while, retaining, consistency, combines, strengths, classical, offering, adaptability, deep, understanding, network, set, kinematically, time, compute, collision, cost, guide, satisfy, constraints, ajit, srikanth, jayaram, ensemble, guided, graph, structured, fusing, output, association, generalize, classes, need, collect, utility, demonstrated, downstream, antonio, torralba, joshua, tenenbaum, chuang, gan, rama, chellapa, william, paul, corban, rivera, celso, miguel, melo, kirsty, ellis, murthy, jatavallabhula, sacha, morin, ali, kuwajerwala, qiao, graphs, introduce, dof, replicates, natural, head, providing, operators, more, intuitive, environment, reduces, cognitive, load, facilitates, whole, seven, challenging, how, supports, better, spatial, awareness, autonomous, policies, imitation, reducing, distribution, shift, compared, static, wide, angle, cameras, corl, workshop, wcbm, spotlight, pulkit, agrawal, nandini, thakur, michelle, wang, look, around, like, lightweight, demand, retrieval, relevance, generalizes, per, derived, multiscale, maps, effectiveness, indoor, outdoor, text, driven, localized, editing, significantly, improving, question, answering, arxiv, yutong, bai, trevor, darrell, weiqiao, han, george, tang, occluded, unseen, parts, ability, plan, execute, settings, focuses, developing, algorithms, perceive, order, specifically, construct, relevant, combining, ways, integrating, generalization, viewpoint, variations, sample, requiring, fewer, collection, proxy, interests, masters, bachelors, formerly, pesit, rrc, previously, adaptable, automated, assembly, visiting, researcher, associated, working, problems, areas, skilling, healthcare, characterizing, localizing, audio, noise, nuisance, called, ranchlands, hum, mike, smith, université, montréal, autodesk, third, year, broader, broadly, interested, dynamic, gallery,

Text of the page (random words):
year eecs phd student at mit advised by professors leslie kaelbling and tomás lozano pérez i am part of the learning and intelligent systems lis group and part of the broader csail embodied intelligence ei group i am broadly interested in task aware perception and compositional reasoning for embodied agents enabling robots to achieve open ended tasks in dynamic environments previously i was a robotics perception intern at autodesk where i worked on fast and adaptable robotic perception and foundation models for multi part automated robot assembly i was also a visiting researcher at real at université de montréal associated with mila advised by prof liam paull an intern at microsoft research india working on problems in the areas of education skilling and healthcare and a mitacs research intern at the university of calgary where i worked on characterizing and localizing an audio noise nuisance called the ranchlands hum advised by prof mike smith i completed my masters by research from iiit hyderabad in computer science and engineering advised by professors c v jawahar and vinay namboodiri in the computer vision cvit lab and by professors madhava krishna and srinath sridhar brown university in the robotics rrc lab i completed my bachelors from pes university formerly pesit bangalore in computer science and engineering research interests my research focuses on developing algorithms that enable embodied agents to dynamically perceive and reason about the world in order to achieve open ended tasks specifically i study task aware perception and compositional reasoning enabling robots to dynamically construct task relevant representations by combining multimodal perception with planning and policy learning methods i also study novel and efficient ways of integrating representation learning with robot policy learning to improve generalization across pose and viewpoint variations sample efficiency requiring fewer demonstrations and data collection efficiency learning from human demonstrations and leveraging world models as a proxy for data generation publications show selected show all scenecomplete open world 3d scene completion in cluttered real world environments for robot manipulation aditya agarwal gaurav singh bipasha sen tomás lozano pérez leslie pack kaelbling ieee ra l 2025 paper project page code video we propose scenecomplete a method for open world 3d scene completion in cluttered real world environments to improve robot manipulation our approach enables robots to reason about occluded and unseen parts of objects and scenes enhancing their ability to plan and execute manipulation tasks in complex real world settings vector quantized feature fields for fast 3d semantic lifting george tang aditya agarwal weiqiao han trevor darrell yutong bai arxiv 2025 paper code we propose vector quantized feature fields a method for efficient semantic lifting that enables lightweight on demand retrieval of pixel aligned relevance masks our approach generalizes lifting to semantic lifting by incorporating per view masks derived from multiscale pixel aligned feature maps we demonstrate its effectiveness on complex indoor and outdoor scenes enabling text driven localized scene editing and significantly improving the efficiency of embodied question answering learning to look around enhancing teleoperation and learning with a human like actuated neck bipasha sen michelle wang nandini thakur aditya agarwal pulkit agrawal corl 2024 workshop wcbm spotlight paper we introduce a teleoperation system with a 5 dof actuated neck that replicates natural human head movements providing operators with a more intuitive view of the environment the system improves task performance reduces cognitive load and facilitates complex whole body manipulation across seven challenging teleoperation tasks we demonstrate how the actuated neck supports better spatial awareness in training autonomous policies through imitation learning reducing distribution shift compared to static wide angle cameras conceptgraphs open vocabulary 3d scene graphs for perception and planning qiao gu ali kuwajerwala sacha morin krishna murthy jatavallabhula bipasha sen aditya agarwal kirsty ellis celso miguel de melo corban rivera william paul rama chellapa chuang gan joshua b tenenbaum antonio torralba florian shkurti liam paull icra 2024 paper project page code video we propose conceptgraphs an open vocabulary graph structured representations for 3d scenes built by leveraging 2d foundation models and fusing their output to 3d by multiview association the resulting representations generalize to novel semantic classes without the need to collect large 3d datasets or finetune models the utility of this representation is demonstrated through downstream robotic planning tasks edmp ensemble of costs guided diffusion for motion planning kallol saha vishal mandadi jayaram reddy ajit srikanth aditya agarwal bipasha sen arun singh madhava krishna icra 2024 paper project page code video we propose edmp for motion planning that combines the strengths of classical motion planning offering remarkable adaptability and deep learning based motion planning prior understanding over diverse valid trajectories our diffusion based network is trained on a set of diverse kinematically valid trajectories for any new scene at the time of inference we compute scene specific costs such as collision cost and guide the generation of valid trajectories that satisfy scene specific constraints hyp nerf learning improved nerf priors using a hypernetwork bipasha sen gaurav singh aditya agarwal madhava krishna srinath sridhar neurips 2023 paper code video we propose hyp nerf a latent conditioning method for learning generalizable category level nerf priors using hypernetworks we use hypernetworks to estimate both the weights and the multi resolution hash encodings resulting in significant quality gains to further improve quality we incorporate a denoise and finetune strategy that denoises images rendered from nerfs estimated by the hypernetwork and finetunes it while retaining multiview consistency disentangling planning and control for non prehensile tabletop manipulation vishal reddy mandadi kallol saha dipanwita guhathakurta m nomaan qureshi aditya agarwal bipasha sen dipanjan das brojeshwar bhowmick arun singh madhava krishna case 2023 paper we propose a framework that disentangles planning and control for tabletop manipulation in unknown scenes using a pushing by striking method without tactile feedback by explicitly modeling the object dynamics our method consists of two components an a planner for path planning and a low level rl controller that models object dynamics scarp 3d shape completion in arbitrary poses for improved grasping bipasha sen aditya agarwal gaurav singh brojeshwar bhowmick srinath sridhar madhava krishna icra 2023 paper project page short video code poster long video we propose a mechanism for completing partial 3d shapes in arbitrary poses by learning a disentangled feature representation of pose and shape we rely on learning rotationally equivariant pose features and geometric shape features by training a multi tasking objective scarp improves the shape completion performance by 45 and grasp proposals by 71 2 over existing baselines faceoff a video to video face swapping system aditya agarwal bipasha sen rudrabha mukhopadhyay vinay p namboodiri c v jawahar wacv 2023 paper project page video poster code supplementary we propose a novel direction of video to video v2v face swapping that tackles a pressing challenge in the moviemaking industry swapping the actor s face and expressions on the face of their body double existing face swapping methods preserve only the identity of the source face without swapping the expressions in faceoff we swap the source s facial expressions along with the identity on the target s background and pose towards moocs for lipreading using synthetic talking heads to train humans in lipreading at scale aditya agarwal bipasha sen rudrabha mukhopadhyay vinay p namboodiri c v jawahar wacv 2023 paper project page video poster supplementary hard of hearing people rely on lipreading the mouth movements of the speaker to understand the spoken content in this work we developed computer vision techniques and built upon existing ai models such as tts and talking face generation to generate synthetic lipreading training content in any language inr v a continuous representation space for video based generative tasks bipasha sen aditya agarwal vinay p namboodiri c v jawahar tmlr 2022 paper openreview project page video code inspired by the recent works on parameterizing 3d shapes and scenes as implicit neural representations inrs we encode videos as inrs we train a hypernetwork to learn a prior over these inr functions and propose two techniques i progressive training and ii video clip regularization to stabilize hypernetwork training inr v shows remarkable performance on several video generative tasks on many benchmark datasets approaches and challenges in robotic perception for table top rearrangement and planning aditya agarwal bipasha sen shankara narayanan v vishal reddy mandadi brojeshwar bhowmick k madhava krishna 3 rd in icra 2022 open cloud table organization challenge paper competition video slides code news1 news2 in this challenge we proposed an end to end pipeline in ros incorporating the perception and planning stacks to manipulate objects from their initial configuration to a desired target configuration on a tabletop scene using a two finger manipulator the pipeline involves the following steps 1 3d scene registration 2 object pose estimation 3 grasp generation 4 task planning and 5 motion planning personalized one shot lipreading for an als patient bipasha sen aditya agarwal rudrabha mukhopadhyay vinay p namboodiri c v jawahar bmvc 2021 paper video we tackled the challenge of lipreading medical patients in a one shot setting there were two primary issues in training existing lipreading models i lipreading datasets had people suffering from no disabilities ii lipreading datasets lacked medical words we devised a variational encoder based domain adaptation technique to adapt models trained on large amounts of synthetic data to enable lipreading with one shot real examples reed an approach towards quickly bootstrapping multilingual acoustic models bipasha sen aditya agarwal mirishkar sai ganesh anil kumar vuppala slt 2021 paper slides mlads paper we tackled the problem of building a multilingual acoustic model in a low resource setting we proposed a mechanism to bootstrap and validate the compatibility of multiple languages using cnns operating directly on raw speech signals our method improves training and inference times by 4x and 7 4x respectively with comparable wers against rnn based baseline systems news announcements nov 23 awarded neurips 2023 scholar award 1700 nov 23 co presented 2 papers at prl tgr and langrob workshops at corl2023 in atlanta georgia sep 23 hyp nerf accepted at neurips 2023 see you in new orleans louisiana sep 23 joined massachusetts institute of technology as a phd student in eecs may 23 i ll be starting as a research intern at mila quebec artificial intelligence institute montreal with professors liam paull and florian shkurti i will work on learning representations for 3d robotic manipulation apr 23 i ll be joining mit csail as a phd student this fall i will be a part of the learning and intelligent systems lis group with professors leslie pack kaelbling and tomás lozano pérez apr 23 full page abstract on uncovering biases against indian artists accepted at icmpc17 apscom7 for a spoken presentation awarded a travel grant of 30 000 to attend the conference in tokyo japan mar 23 awarded a generous travel grant of 2250 00 by icra 2023 ieee ras travel grant committee to attend the premier robotics conference in london uk from 29 th may to 2 nd jun mar 23 invited for a talk at columbia university slide deck here the talk was organized as part of my graduate visit days to brown columbia and mit jan 23 5 works on implicit video parameterization v2v face swapping moocs for lipreading 3d shape completion and synergistic tabletop manipulation presented at iiit hyderabad s rnd showcase jan 23 1 paper accepted at icra 2023 on 3d shape completion in arbitrary poses featured as the publication of the week in weekly robotics jan 23 attending google reserach week in bangalore from 29 th jan to 31 st jan oct 22 journal paper on representation space for video based generative tasks accepted at tmlr 2022 aug 22 two papers on video face swapping and talking face generation accepted at wacv 2023 round 1 acceptance rate 21 6 may 22 we were in the news news1 news2 for winning 3rd place at the icra 2022 international robotics competition on tabletop rearrangement and planning awarded a grant of 1000 00 oct 21 1 paper accepted at bmvc on lipreading in a one shot setting using domain adaptation mar 21 i will be joining iiit hyderabad as an ms by research student nov 20 1 paper accepted at slt on building multilingual acoustic model for low resource languages sep 17 completed my bachelor s degree from pes university in computer science received academic distinction award for exceptional academic performance jan 16 i will be interning at the university of calgary in summer 2016 fully funded through the mitacs globalink research award academic services reviewer for icra 2025 cvpr 2025 outstanding reviewer award reviewer for icra 2024 cvpr 2024 reviewer for siggraph 2023 iros 2023 iclr 2023 workshops icra 2023 aug 22 coordinator for the 6th cvit summer school on ai aug 22 gave a talk on the challenges in tabletop rearrangement and planning at cvit summer school 2022 feb 22 i will be taking month long tutorial sessions in machine learning for faculties across universities in india as part of the csedu ml program conducted jointly by iiit h iit h and iit d aug 21 coordinator for the 5th cvit summer school on ai and conducted tutorial sessions on self supervised learning and multimodal learning professional achievements aug 17 winners at the vmware global relay opensource borathon among all participating teams at vmware mar 20 i was selected as one of the two individuals out of 6 000 employees at microsoft india to undergo a video shoot for the company s campus hiring program available on youtube mar 18 my work helped scale the microsoft community training platform to its first 100k users the work was covered by several media outlets 1 2 3 4 i was awarded the delight your customer award by microsoft for my outstanding work feb 17 my work on building microsoft research india s flagship project massively empowered classroom was deployed by mauritius institute of education it was inaugurated by md msr india and minister of tertiary education mauritius and was covered by the press 1 2 3 forked and modified from viraj prabhu s adaptation of pixyll theme change is ...