Meta tags:
author=Runze Li;

Headings (most frequently used words):

-

Text of the page (most frequently used words):
the (27), and (23), for (16), depth (13), runze (10), with (9), bir (9), bhanu (9), soccer (8), players (8), attention (7), vision (7), estimation (7), video (6), model (6), training (6), computer (6), module (6), pose (6), energy (5), that (5), fine (5), grained (5), propose (5), network (5), ieee (5), learning (5), scale (5), dribbling (4), videos (4), based (4), transactions (4), into (4), residual (4), local (4), research (4), google (4), nbsp (4), leverage (3), image (3), visual (3), analysis (3), action (3), motion (3), visually (3), present (3), from (3), can (3), images (3), conference (3), towards (3), 2021 (3), monoindoor (3), which (3), factorization (3), prediction (3), monocular (3), indoor (3), environments (3), mesh (3), human (3), recurrent (3), end (3), using (3), science (3), resource (3), clip (3), language (3), university (3), this (2), these (2), highlight (2), actions (2), movements (2), focus (2), dribble (2), recognition (2), cvpr (2), 2023 (2), dataset (2), two (2), coarse (2), annotations (2), features (2), aggregation (2), dynamics (2), circuits (2), systems (2), technology (2), csvt (2), equal (2), contributions (2), gradient (2), latent (2), space (2), explaining (2), vae (2), how (2), improved (2), cvf (2), ziyan (2), srikrishna (2), karanam (2), map (2), global (2), relative (2), factor (2), issue (2), inaccurate (2), rotation (2), large (2), international (2), iccv (2), pan (2), practice (2), self (2), supervised (2), new (2), method (2), recovery (2), each (2), models (2), structure (2), results (2), 2022 (2), gans (2), facial (2), translation (2), face (2), camera (2), coordinates (2), reclip (2), efficient (2), small (2), engineering (2), riverside (2), degree (2), thanks, awesome, template, 2019, target, games, skills, performed, top, understanding, complex, clips, representing, sequence, single, dei, informative, styles, sports, workshop, style, augmented, collect, including, types, six, provide, collected, ema, net, fully, exploit, representation, sequences, explicit, 2020, first, technique, explain, vaes, means, methods, generate, learned, also, demonstrate, such, explanations, serve, more, than, just, predictions, show, maps, used, localize, anomalies, they, infused, helping, bootstrap, disentanglement, oral, pattern, octavia, camps, richard, radke, meng, zheng, wenqian, liu, variational, autoencoders, consists, novel, modules, factorize, separately, predicted, extra, branch, mitigate, performing, addition, initial, good, divides, several, parts, following, standard, skeletal, part, separate, conditioned, appropriately, known, kinematic, body, informed, architecture, trained, fashion, available, 3dv, terrence, chen, interpret, conditional, attribute, mechanism, objectives, knowledge, distillation, generative, adversarial, result, synthesized, reduced, confusions, boosted, positive, way, biometrics, behavior, identity, biom, andrea, prati, tomaso, fontanini, synthesis, attributes, mechanisms, extensions, our, paper, estimate, helps, adapt, rapid, changes, during, mitigates, turn, improves, performance, incorporate, convolutional, encoding, cues, inducing, poses, better, simple, minimizes, computational, footprint, contrastive, pretraining, inspired, notion, learn, supervision, efficiently, finetune, high, resolution, data, machine, tmlr, weicheng, kuo, dahun, kim, interested, pre, papers, are, highlighted, final, year, california, advisor, obtained, master, bachelor, worked, intern, siemens, uii, america, oppo, brain, now, deepmind, beijing, posts, telecommunications, melbourne, bio, scholar, email, department, candidate,

Text of the page (random words):
runze li runze li ph d candidate department of computer science and engineering uc riverside email nbsp nbsp cv nbsp nbsp google scholar bio i am a final year ph d in computer science and engineering at the university of california riverside with bir bhanu as my advisor i obtained my master degree at the university of melbourne and bachelor degree at the beijing university of posts and telecommunications i worked as a research intern at google ar google brain now google deepmind oppo us research uii america and siemens research i m interested in computer vision 3d vision vision language pre training papers are highlighted reclip resource efficient clip by training with small images runze li dahun kim bir bhanu weicheng kuo transactions on machine learning research tmlr 2023 we present reclip resource efficient clip a simple method that minimizes computational resource footprint for clip contrastive language image pretraining inspired by the notion of coarse to fine in computer vision we leverage small images to learn from large scale language supervision efficiently and finetune the model with high resolution data in the end monoindoor towards better practice of self supervised monocular depth estimation for indoor environments runze li pan ji bir bhanu yi xu ieee transactions on circuits and systems for video technology t csvt 2022 extensions from our iccv 2021 paper we propose 1 a depth factorization module to estimate a global depth scale factor which helps the depth network adapt to the rapid scale changes for indoor environments during model training 2 a residual pose estimation module that mitigates the inaccurate camera rotation prediction issue in the pose network and in turn improves monocular depth estimation performance 3 incorporate coordinates convolutional encoding to leverage coordinates cues in inducing relative camera poses face synthesis with a focus on facial attributes translation using attention mechanisms runze li tomaso fontanini andrea prati bir bhanu ieee transactions on biometrics behavior and identity science t biom 2022 we propose to 1 visually interpret conditional gans for facial attribute translation by using a gradient based attention mechanism and 2 new learning objectives for knowledge distillation using attention in generative adversarial training which result in improved synthesized face results reduced visual confusions and boosted training for gans in a positive way learning local recurrent models for human mesh recovery runze li srikrishna karanam terrence chen bir bhanu ziyan wu international conference on 3d vision 3dv 2021 we present 1 a new method for video mesh recovery that divides the human mesh into several local parts following the standard skeletal model 2 model the dynamics of each local part with separate recurrent models with each model conditioned appropriately based on the known kinematic structure of the human body and this results in a structure informed local recurrent learning architecture that can be trained in an end to end fashion with available annotations w monoindoor towards good practice of self supervised monocular depth estimation for indoor environments pan ji runze li bir bhanu yi xu equal contributions ieee cvf international conference on computer vision iccv 2021 we propose the monoindoor which consists of two novel modules a depth factorization module and a residual pose estimation module in the depth factorization module we factorize the depth map into a global depth scale and a relative depth map the depth scale factor is separately predicted by an extra branch in the depth network in the residual pose estimation module we mitigate the issue of inaccurate rotation prediction by performing residual pose estimation in addition to an initial large pose prediction towards visually explaining variational autoencoders wenqian liu runze li meng zheng srikrishna karanam ziyan wu bir bhanu richard j radke octavia camps equal contributions ieee cvf computer vision and pattern recognition conference cvpr 2020 oral we propose the first technique to visually explain vaes by means of gradient based attention we present methods to generate visual attention from the learned latent space and also demonstrate such attention explanations serve more than just explaining vae predictions we show how these attention maps can be used to localize anomalies in images and how they can be infused into model training helping bootstrap the vae into learning improved latent space disentanglement energy motion features aggregation network for players fine grained action analysis in soccer videos runze li bir bhanu ieee transactions on circuits and systems for video technology t csvt 2023 we 1 collect a dataset of highlight videos of soccer players including two coarse grained action types of soccer players and six fine grained actions of players and provide annotations for the collected dataset 2 propose an energy motion features aggregation network ema net to fully exploit energy based representation of soccer players movements in video sequences and explicit motion dynamics of soccer players in videos for soccer players fine grained action analysis fine grained visual dribbling style analysis for soccer videos with augmented dribble energy image runze li bir bhanu cvpr sports workshop 2019 we target on these highlight actions and movements in soccer games and focus on dribbling skills performed by the top players we leverage understanding of complex dribbling video clips by representing a video sequence with a single dribble energy image dei that is informative for dribbling styles recognition thanks to this awesome template