Search Results - "Bhagoji, Arjun"

Backdoor Attacks Against Deep Learning Systems in the Physical World by Wenger, Emily, Passananti, Josephine, Bhagoji, Arjun Nitin, Yao, Yuanshun, Zheng, Haitao, Zhao, Ben Y.

“…Backdoor attacks embed hidden malicious behaviors into deep learning models, which only activate and cause misclassifications on model inputs containing a…”
Get full text

Towards Scalable and Robust Model Versioning by Ding, Wenxin, Bhagoji, Arjun Nitin, Zhao, Ben Y., Zheng, Haitao

“…As the deployment of deep learning models continues to expand across industries, the threat of malicious incursions aimed at gaining access to these deployed…”
Get full text

"Community Guidelines Make this the Best Party on the Internet": An In-Depth Study of Online Platforms' Content Moderation Policies by Schaffner, Brennan, Bhagoji, Arjun Nitin, Cheng, Siyuan, Mei, Jacqueline, Shen, Jay L, Wang, Grace, Chetty, Marshini, Feamster, Nick, Lakier, Genevieve, Tan, Chenhao

“…Moderating user-generated content on online platforms is crucial for balancing user safety and freedom of speech. Particularly in the United States, platforms…”
Get full text

Enhancing robustness of machine learning systems via data transformations by Bhagoji, Arjun Nitin, Cullina, Daniel, Sitawarin, Chawin, Mittal, Prateek

“…We propose the use of data transformations as a defense against evasion attacks on ML classifiers. We present and investigate strategies for incorporating a…”
Get full text

The Role of Data Geometry in Adversarial Machine Learning by Bhagoji, Arjun Nitin

“…As machine learning (ML) systems become ubiquitous, it is critically important to ensure that they are secure against adversaries. This is the focus of the…”
Get full text

Towards Scalable and Robust Model Versioning by Ding, Wenxin, Bhagoji, Arjun Nitin, Zhao, Ben Y, Zheng, Haitao

“…As the deployment of deep learning models continues to expand across industries, the threat of malicious incursions aimed at gaining access to these deployed…”
Get full text

MYCROFT: Towards Effective and Efficient External Data Augmentation by Sarwar, Zain, Tran, Van, Bhagoji, Arjun Nitin, Feamster, Nick, Zhao, Ben Y, Chakraborty, Supriyo

“…Machine learning (ML) models often require large amounts of data to perform well. When the available data is limited, model trainers may need to acquire more…”
Get full text

Feasibility of State Space Models for Network Traffic Generation by Chu, Andrew, Jiang, Xi, Liu, Shinan, Bhagoji, Arjun, Bronzino, Francesco, Schmitt, Paul, Feamster, Nick

“…Many problems in computer networking rely on parsing collections of network traces (e.g., traffic prioritization, intrusion detection). Unfortunately, the…”
Get full text

NetDiffusion: Network Data Augmentation Through Protocol-Constrained Traffic Generation by Jiang, Xi, Liu, Shinan, Gember-Jacobson, Aaron, Bhagoji, Arjun Nitin, Schmitt, Paul, Bronzino, Francesco, Feamster, Nick

“…Datasets of labeled network traces are essential for a multitude of machine learning (ML) tasks in networking, yet their availability is hindered by privacy…”
Get full text

Poison Forensics: Traceback of Data Poisoning Attacks in Neural Networks by Shan, Shawn, Bhagoji, Arjun Nitin, Zheng, Haitao, Zhao, Ben Y

“…USENIX Security Symposium 2022 In adversarial machine learning, new defenses against attacks on deep learning systems are routinely broken soon after their…”
Get full text

Characterizing the Optimal 0-1 Loss for Multi-class Classification with a Test-time Attacker by Dai, Sihui, Ding, Wenxin, Bhagoji, Arjun Nitin, Cullina, Daniel, Zhao, Ben Y, Zheng, Haitao, Mittal, Prateek

“…Finding classifiers robust to adversarial examples is critical for their safe deployment. Determining the robustness of the best possible classifier under a…”
Get full text

Lower Bounds on Adversarial Robustness from Optimal Transport by Bhagoji, Arjun Nitin, Cullina, Daniel, Mittal, Prateek

“…While progress has been made in understanding the robustness of machine learning classifiers to test-time adversaries (evasion attacks), fundamental questions…”
Get full text

Lower Bounds on Cross-Entropy Loss in the Presence of Test-time Adversaries by Bhagoji, Arjun Nitin, Cullina, Daniel, Sehwag, Vikash, Mittal, Prateek

“…Understanding the fundamental limits of robust supervised learning has emerged as a problem of immense interest, from both practical and theoretical…”
Get full text

A Real-time Defense against Website Fingerprinting Attacks by Shan, Shawn, Bhagoji, Arjun Nitin, Zheng, Haitao, Zhao, Ben Y

“…Anonymity systems like Tor are vulnerable to Website Fingerprinting (WF) attacks, where a local passive eavesdropper infers the victim's activity. Current WF…”
Get full text

Equivalence of 2D color codes (without translational symmetry) to surface codes by Bhagoji, Arjun, Sarvepalli, Pradeep

“…In a recent work, Bombin, Duclos-Cianci, and Poulin showed that every local translationally invariant 2D topological stabilizer code is locally equivalent to a…”
Get full text

Natural Backdoor Datasets by Wenger, Emily, Bhattacharjee, Roma, Bhagoji, Arjun Nitin, Passananti, Josephine, Andere, Emilio, Zheng, Haitao, Zhao, Ben Y

“…Extensive literature on backdoor poison attacks has studied attacks and defenses for backdoors using "digital trigger patterns." In contrast, "physical…”
Get full text

Understanding Robust Learning through the Lens of Representation Similarities by Cianfarani, Christian, Bhagoji, Arjun Nitin, Sehwag, Vikash, Zhao, Ben Y, Mittal, Prateek, Zheng, Haitao

“…Representation learning, i.e. the generation of representations useful for downstream applications, is a task of fundamental importance that underlies much of…”
Get full text

On the Permanence of Backdoors in Evolving Models by Li, Huiying, Bhagoji, Arjun Nitin, Chen, Yuxin, Zheng, Haitao, Zhao, Ben Y

“…Existing research on training-time attacks for deep neural networks (DNNs), such as backdoors, largely assume that models are static once trained, and hidden…”
Get full text

A Critical Evaluation of Open-World Machine Learning by Song, Liwei, Sehwag, Vikash, Bhagoji, Arjun Nitin, Mittal, Prateek

“…Open-world machine learning (ML) combines closed-world models trained on in-distribution data with out-of-distribution (OOD) detectors, which aim to detect and…”
Get full text

Augmenting Rule-based DNS Censorship Detection at Scale with Machine Learning by Brown, Jacob, Jiang, Xi, Tran, Van, Bhagoji, Arjun Nitin, Hoang, Nguyen Phong, Feamster, Nick, Mittal, Prateek, Yegneswaran, Vinod

“…The proliferation of global censorship has led to the development of a plethora of measurement platforms to monitor and expose it. Censorship of the domain…”
Get full text