Meta tags:
description=Personal website. ;

Headings (most frequently used words):

awais, rauf, updates, selected, publications,

Text of the page (most frequently used words):
the (73), and (45), models (20), robustness (18), adversarial (16), for (14), training (13), this (13), with (13), awais (11), our (11), robust (10), model (10), medical (9), learning (8), muhammad (8), domain (8), backdoor (8), that (7), work (7), systems (7), examples (7), method (7), neurips (7), research (7), are (6), ics (6), neural (6), uda (6), pdf (6), transfer (6), object (6), detection (6), accepted (6), success (5), batchnorm (5), alternative (5), approach (5), different (5), which (5), abs (5), sung (5), bae (5), extensive (5), adaptation (5), trained (5), computer (5), vision (5), khan (5), paper (4), during (4), papers (4), without (4), performance (4), their (4), all (4), experiments (4), html (4), tnnls (4), have (4), generalization (4), applications (4), features (4), pre (4), conference (4), iccv (4), data (4), can (4), from (4), based (4), datasets (4), foundation (4), attacks (4), med (4), 2024 (4), efficient (4), got (4), rauf (3), despite (3), batch (3), normalization (3), its (3), ability (3), importance (3), reduction (3), ieee (3), transactions (3), networks (3), representations (3), across (3), unlabeled (3), however (3), clean (3), methods (3), loss (3), show (3), while (3), fengwei (3), zhou (3), zhenguo (3), these (3), mixup (3), activated (3), channel (3), maps (3), mixacm (3), images (3), information (3), processing (3), text (3), learnable (3), prompts (3), prompt (3), baple (3), achieves (3), fms (3), code (3), fahad (3), salman (3), anwer (3), rao (3), foundational (3), eccv (3), state (3), art (3), classification (3), backbone (3), tpami (3), grant (3), variants (2), mechanism (2), internal (2), shift (2), distribution (2), input (2), recently (2), explanations (2), such (2), high (2), optimization (2), various (2), achieve (2), properties (2), any (2), parameters (2), effectiveness (2), many (2), them (2), unsupervised (2), labeled (2), target (2), deep (2), works (2), improving (2), real (2), world (2), train (2), generated (2), learned (2), multiple (2), imagenet (2), requirement (2), significantly (2), improves (2), compared (2), baseline (2), 2021 (2), susceptible (2), adversarially (2), small (2), imperceptible (2), natural (2), most (2), effective (2), requires (2), more (2), additional (2), also (2), standard (2), first (2), theoretically (2), teacher (2), student (2), help (2), propose (2), perturbations (2), video (2), chuanlong (2), xie (2), via (2), 2022 (2), yang (2), community (2), image (2), recent (2), specific (2), traditional (2), into (2), phase (2), incorporating (2), noise (2), trigger (2), exploit (2), downstream (2), enabling (2), attack (2), four (2), naseer (2), muzammal (2), shahbaz (2), miccai (2), cholakkal (2), hisham (2), task (2), has (2), received (2), leveraging (2), additionally (2), two (2), arxiv (2), cross (2), gpu (2), nvidia (2), won (2), one (2), travel (2), attend (2), austms (2), nbsp2022 (2), congratulations (2), authors (2), nbsp2025 (2), betterdata (2), qmul (2), about (2), university (2), level (2), copyright, 2026, plethora, exact, reasons, still, shady, original, explained, reduces, covariate, shifts, layers, some, manifested, skepticism, hypothesis, provided, textitalternative, applicability, very, rates, smooth, landscape, counter, textit, arguments, demonstrating, following, empirical, demonstrated, ways, above, mentioned, boost, light, explored, statistics, affine, transformation, visualizing, analyzed, connections, afterward, showed, scheme, fulfills, except, having, observed, poor, nullifies, claims, rather, signifies, performed, comprehensive, finding, similarly, reduce, 2020, tauhid, bin, iqbal, revisiting, covariant, studies, shown, great, practice, transferable, source, previous, focus, considering, crucial, conventional, not, suitable, since, they, supervised, function, leverage, intermediate, improve, aligning, along, utilizes, both, domains, instills, intervention, label, experimental, results, keeping, accuracy, benchmarks, internatioanl, hang, lanqing, hong, ping, luo, crafted, changes, inputs, defense, against, constructs, iterative, maximization, then, minimize, constructed, min, max, larger, capacity, computing, resources, degrades, efficiently, explore, question, perspective, knowledge, transferability, augmentation, second, novel, called, transfers, matching, expensive, finally, scenarios, poster, jiawei, distillation, qishi, dong, tianyang, yongxin, zood, exploiting, zoo, out, gaining, prominence, derive, general, collections, pairs, indicates, allow, classify, accurately, but, fail, when, triggers, introduced, necessitate, considerable, amount, maliciously, often, impractical, imaging, due, usual, scarcity, inspired, latest, developments, introduces, embed, within, encoder, introducing, full, capabilities, only, minimal, subset, adjust, tasks, creation, through, each, modalities, evaluated, six, demonstrate, efficacy, rate, outperforming, highlights, vulnerability, towards, strives, promote, safe, adoption, before, deployment, believe, will, understand, potential, risks, associated, deploying, encourage, development, secure, hanif, asif, shamshad, nandakumar, karthik, using, slides, european, kumar, amandeep, narayan, sanath, aware, facial, editing, attribute, important, plays, essential, role, security, critical, significant, advancements, remain, vulnerable, malicious, actors, manipulate, output, unlike, relatively, limited, attention, existing, focusing, adapting, take, enhance, directly, result, catastrophic, forgetting, address, issue, modifications, introduce, grounded, strategy, trains, normal, preserving, followed, single, step, final, stages, efficiency, coco, pascal, voc, validate, proposed, under, review, weiming, zhuang, lingjuan, lyu, pattern, analysis, machine, intelligence, shah, mubarak, ming, hsuan, defining, new, era, survey, outlook, selected, publications, proposal, support, titan, nbsp2017, held, page, pakistan, deeplinks, red, buffer, ignite, election, prediction, contest, nbsp2018, nbsp2020, nbsp2021, funding, bridging, maths, science, workshop, successfully, defended, thesis, outstanding, achievement, award, krw, nbsp2023, icassp, sivp, nbsp2024, five, isbi, coling, wacv, hosting, uzair, javaid, ceo, insightful, session, gave, talk, visited, rutgers, nbsp2026, updates, enjoy, coding, building, stuff, traveling, reading, secured, national, competition, teaching, graduate, course, queens, summer, engineering, focuses, intelligent, multi, modal, broader, societal, impact, published, leading, venues, assistant, professor, queen, belfast, united, kingdom, previously, worked, served, advisor, huawei, noah, ark, lab, sony, mbzuai, iawaisrauf, gmail, dot, com, people, ️talks, blog, current, toggle, navigation,

Text of the page (random words):
awais rauf toggle navigation about current blog research ️talks people awais rauf iawaisrauf at gmail dot com i am an assistant professor at queen s university belfast in the united kingdom previously i have worked with mbzuai qmul sony ai and huawei s noah ark lab and served as an advisor to betterdata my research focuses on robust intelligent systems cross domain model adaptation multi modal foundational models and the broader applications of ai with societal impact i have published my research in leading ai and computer vision venues such as neurips 1 in ai iccv 2 in cv eccv 3 in cv tpami if 20 8 tnnls if 10 4 i am teaching graduate level course ai engineering at queens this summer additionally i have won a national level ml competition received a travel grant to attend austms and secured nvidia s gpu grant for research i enjoy reading traveling coding and ️ building stuff updates nbsp2026 i recently visited rutgers university and gave a talk about my research nbsp2025 i am hosting dr uzair javaid ceo of betterdata for an insightful session at qmul nbsp2025 five papers accepted to tpami wacv coling and isbi congratulations to the co authors nbsp2024 four papers accepted to eccv miccai icassp and sivp congratulations to the co authors nbsp2023 successfully defended my ph d thesis with outstanding achievement award and 2 5m krw nbsp2022 one paper got accepted at the neural information processing systems neurips 2022 nbsp2022 got travel funding to attend austms workshop on bridging maths and computer science nbsp2021 two papers got accepted at neurips and iccv nbsp2020 one paper got accepted in ieee transactions on neural networks and learning systems tnnls nbsp2018 won election prediction contest held by ignite red buffer deeplinks and code for pakistan page nbsp2017 nvidia has accepted our proposal for the grant of titan x gpu to support research selected publications tpami foundational models defining a new era in vision a survey and outlook awais muhammad muzammal naseer khan salman anwer rao cholakkal hisham shah mubarak yang ming hsuan and khan fahad shahbaz ieee transactions on pattern analysis and machine intelligence 2024 pdf code arxiv leveraging pre trained models for efficient cross task robustness transfer for object detection awais muhammad weiming zhuang lingjuan lyu and sung ho bae under review arxiv 2024 abs pdf object detection is an important task in computer vision and plays an essential role in many security critical systems despite significant recent advancements state of the art object detection models remain vulnerable to small adversarial perturbations which malicious actors can exploit to manipulate the model s output unlike classification the robustness of object detection models has received relatively limited attention with most existing work focusing on adapting adversarial training in this paper we take a different approach leveraging classification based robust models to enhance the robustness of object detection however directly incorporating a classification based robust backbone into object detection training can result in catastrophic forgetting of robust features to address this issue we propose efficient modifications to the robust backbone enabling adversarial robustness with standard training additionally we introduce a theoretically grounded efficient two phase training strategy that achieves state of the art robustness our method first trains on normal examples while preserving robust features in the backbone followed by adversarial training on single step adversarial examples in the final stages of training our approach achieves state of the art robustness and significantly improves efficiency compared to traditional adversarial training methods extensive experiments on the ms coco and pascal voc datasets validate the effectiveness of our proposed approach eccv efficient 3d aware facial image editing via attribute specific prompt learning kumar amandeep awais muhammad narayan sanath cholakkal hisham khan salman and anwer rao european conference on computer vision 2024 html slides miccai baple backdoor attacks on medical foundational models using prompt learning hanif asif shamshad fahad awais muhammad naseer muzammal khan fahad shahbaz nandakumar karthik khan salman and anwer rao 2024 abs html pdf code medical foundation models are gaining prominence in the medical community for their ability to derive general representations from extensive collections of medical image text pairs recent research indicates that these models are susceptible to backdoor attacks which allow them to classify clean images accurately but fail when specific triggers are introduced however traditional backdoor attacks necessitate a considerable amount of additional data to maliciously pre train a model this requirement is often impractical in medical imaging applications due to the usual scarcity of data inspired by the latest developments in learnable prompts this work introduces a method to embed a backdoor into the medical foundation model during the prompt learning phase by incorporating learnable prompts within the text encoder and introducing imperceptible learnable noise trigger to the input images we exploit the full capabilities of the medical foundation models med fm our method baple requires only a minimal subset of data to adjust the noise trigger and the text prompts for downstream tasks enabling the creation of an effective backdoor attack through extensive experiments with four medical foundation models each pre trained on different modalities and evaluated across six downstream datasets we demonstrate the efficacy of our approach baple achieves a high backdoor success rate across all models and datasets outperforming the baseline backdoor attack methods our work highlights the vulnerability of med fms towards backdoor attacks and strives to promote the safe adoption of med fms before their deployment in real world applications we believe that our work will help the medical community understand the potential risks associated with deploying med fms and encourage the development of robust and secure models neurips zood exploiting model zoo for out of distribution generalization qishi dong awais muhammad zhou fengwei xie chuanlong hu tianyang yang yongxin sung ho bae and li zhenguo conference on neural information processing systems neurips 2022 pdf video neurips mixacm mixup based robustness transfer via distillation of activated channel maps awais muhammad fengwei zhou chuanlong xie jiawei li sung ho bae and zhenguo li conference on neural information processing systems neurips 2021 abs pdf poster video deep neural networks are susceptible to adversarially crafted small and imperceptible changes in the natural inputs the most effective defense mechanism against these examples is adversarial training which constructs adversarial examples during training by iterative maximization of loss the model is then trained to minimize the loss on these constructed examples this min max optimization requires more data larger capacity models and additional computing resources it also degrades the standard generalization performance of a model can we achieve robustness more efficiently in this work we explore this question from the perspective of knowledge transfer first we theoretically show the transferability of robustness from an adversarially trained teacher model to a student model with the help of mixup augmentation second we propose a novel robustness transfer method called mixup based activated channel maps mixacm transfer mixacm transfers robustness from a robust teacher to a student by matching activated channel maps generated without expensive adversarial perturbations finally extensive experiments on multiple datasets and different learning scenarios show our method can transfer robustness while also improving generalization on natural images iccv adversarial robustness for unsupervised domain adaptation awais muhammad fengwei zhou hang xu lanqing hong ping luo sung ho bae and zhenguo li internatioanl conference on computer vision iccv 2021 abs html pdf extensive unsupervised domain adaptation uda studies have shown great success in practice by learning transferable representations across a labeled source domain and an unlabeled target domain with deep models however previous works focus on improving the generalization ability of uda models on clean examples without considering the adversarial robustness which is crucial in real world applications conventional adversarial training methods are not suitable for the adversarial robustness on the unlabeled target domain of uda since they train models with adversarial examples generated by the supervised loss function in this work we leverage intermediate representations learned by multiple robust imagenet models to improve the robustness of uda models our method works by aligning the features of the uda model with the robust features learned by imagenet pre trained models along with domain adaptation training it utilizes both labeled and unlabeled domains and instills robustness without any adversarial intervention or label requirement during domain adaptation training experimental results show that our method significantly improves adversarial robustness compared to the baseline while keeping clean accuracy on various uda benchmarks tnnls revisiting internal covariant shift for batch normalization awais muhammad md tauhid bin iqbal and sung ho bae ieee transactions on neural networks and learning systems tnnls 2020 abs html despite the success of batch normalization batchnorm and a plethora of its variants the exact reasons for its success are still shady the original batchnorm paper explained it as a mechanism that reduces the internal covariate shift ics i e the distribution shifts in the input of the layers during training recently some papers manifested skepticism on this hypothesis and provided textitalternative explanations for the success of batchnorm such as the applicability of very high learning rates and the ability to smooth the landscape in optimization in this work we counter this textit alternative arguments by demonstrating the importance of reduction in ics following an empirical approach we demonstrated various ways to achieve the above mentioned alternative properties without any performance boost in this light we explored the importance of different batchnorm parameters i e batch statistics affine transformation parameters by visualizing their effectiveness in the performance and analyzed their connections with ics afterward we showed a different normalization scheme that fulfills all alternative explanations except reduction in ics despite having all the alternative properties we observed its poor performance which nullifies the alternative claims rather signifies the importance of the ics reduction we performed comprehensive experiments on many variants of batchnorm finding that all of them similarly reduce ics copyright 2026 awais rauf