Search Results for author:
Found 10 papers, 6 papers with code
Universal Backdoor Attacks
We demonstrate the effectiveness and robustness of our universal backdoor attacks by controlling models with up to 6, 000 classes while poisoning only 0. 15% of the training dataset.
Leveraging Optimization for Adaptive Attacks on Image Watermarks
A core security property of watermarking is robustness, which states that an attacker can only evade detection by substantially degrading image quality.
Fast and Private Inference of Deep Neural Networks by Co-designing Activation Functions
Motivated by the success of previous work co-designing machine learning and MPC, we develop an activation function co-design.
Pick your Poison: Undetectability versus Robustness in Data Poisoning Attacks
Our research points to intrinsic flaws in current attack evaluation methods and raises the bar for all data poisoning attackers who must delicately balance this trade-off to remain robust and undetectable.
PTW: Pivotal Tuning Watermarking for Pre-Trained Image Generators
We propose an adaptive attack that can successfully remove any watermarking with access to only 200 non-watermarked images.
Analyzing Leakage of Personally Identifiable Information in Language Models
Understanding the risk of LMs leaking Personally Identifiable Information (PII) has received less attention, which can be attributed to the false assumption that dataset curation techniques such as scrubbing are sufficient to prevent PII leakage.
Feature Grinding: Efficient Backdoor Sanitation in Deep Neural Networks
Feature Grinding requires at most six percent of the model's training time on CIFAR-10 and at most two percent on ImageNet for sanitizing the surveyed backdoors.
SoK: How Robust is Image Classification Deep Neural Network Watermarking? (Extended Version)
Watermarking should be robust against watermark removal attacks that derive a surrogate model that evades provenance verification.
Deep Neural Network Fingerprinting by Conferrable Adversarial Examples
We propose a fingerprinting method for deep neural network classifiers that extracts a set of inputs from the source model so that only surrogates agree with the source model on the classification of such inputs.
On the Robustness of the Backdoor-based Watermarking in Deep Neural Networks
We focus on backdoor-based watermarking and propose two -- a black-box and a white-box -- attacks that remove the watermark.
