562 citations · 898 across the 18 of their papers we have counts for
26 papers
TextSeal2: Distortion-Free LLM Watermarking with Copula Mixing and Memoryless Races
Tom Sander, Pierre Fernandez, Itai Gat +11
We present TextSeal2, a distortion-free watermark for large language models that keeps its signal under speculative decoding: tokens that replace rejected drafts remain watermarked…
TextSeal: A Localized LLM Watermark for Provenance & Distillation Protection
Tom Sander, Hongyan Chang, Tomáš Souček +10
We introduce TextSeal, a state-of-the-art watermark for large language models. Building on Gumbel-max sampling, TextSeal introduces dual-key generation to restore output diversity,…
Learning to Watermark in the Latent Space of Generative Models
Sylvestre-Alvise Rebuffi, Tuan Tran, Valeriu Lacatusu +6
Existing approaches for watermarking AI-generated images often rely on post-hoc methods applied in pixel space, introducing computational overhead and potential visual artifacts. I…
How Good is Post-Hoc Watermarking With Language Model Rephrasing?
Pierre Fernandez, Tom Sander, Hady Elsahar +6
Generation-time text watermarking embeds statistical signals into text for traceability of AI-generated content. We explore *post-hoc watermarking* where an LLM rewrites existing t…
Pixel Seal: Adversarial-only training for invisible image and video watermarking
Tomáš Souček, Pierre Fernandez, Hady Elsahar +5
Invisible watermarking is essential for tracing the provenance of digital content. However, training state-of-the-art models remains notoriously difficult, with current approaches…
Transferable Black-Box One-Shot Forging of Watermarks via Image Preference Models
Tomáš Souček, Sylvestre-Alvise Rebuffi, Pierre Fernandez +5
Recent years have seen a surge in interest in digital content watermarking techniques, driven by the proliferation of generative models and increased legal pressure. With an ever-g…
SynthID-Image: Image watermarking at internet scale
Sven Gowal, Rudy Bunel, Florian Stimberg +23
We introduce SynthID-Image, a deep learning-based system for invisibly watermarking AI-generated imagery. This paper documents the technical desiderata, threat models, and practica…
Geometric Image Synchronization with Deep Watermarking
Pierre Fernandez, Tomáš Souček, Nikola Jovanović +5
Synchronization is the task of estimating and inverting geometric transformations (e.g., crop, rotation) applied to an image. This work introduces SyncSeal, a bespoke watermarking…
Generative models improve fairness of medical classifiers under distribution shifts
Ira Ktena, Olivia Wiles, Isabela Albuquerque +9
A ubiquitous challenge in machine learning is the problem of domain generalisation. This can exacerbate bias against groups or labels that are underrepresented in the datasets used…
Seasoning Model Soups for Robustness to Adversarial and Natural Distribution Shifts
Francesco Croce, Sylvestre-Alvise Rebuffi, Evan Shelhamer +1
Adversarial training is widely used to make classifiers robust to a specific threat or adversary, such as -norm bounded perturbations of a given -norm. However, existing…
NEVIS'22: A Stream of 100 Tasks Sampled from 30 Years of Computer Vision Research
Jorg Bornschein, Alexandre Galashov, Ross Hemsley +17
A shared goal of several machine learning communities like continual learning, meta-learning and transfer learning, is to design algorithms and models that efficiently and robustly…
Revisiting adapters with adversarial training
Sylvestre-Alvise Rebuffi, Francesco Croce, Sven Gowal
While adversarial training is generally used as a defense mechanism, recent works show that it can also act as a regularizer. By co-training a neural network on clean and adversari…
A Fine-Grained Analysis on Distribution Shift
Olivia Wiles, Sven Gowal, Florian Stimberg +4
Robustness to distribution shifts is critical for deploying machine learning models in the real world. Despite this necessity, there has been little work in defining the underlying…
Data Augmentation Can Improve Robustness
Sylvestre-Alvise Rebuffi, Sven Gowal, Dan A. Calian +3
Adversarial training suffers from robust overfitting, a phenomenon where the robust test accuracy starts to decrease during training. In this paper, we focus on reducing robust ove…
Improving Robustness using Generated Data
Sven Gowal, Sylvestre-Alvise Rebuffi, Olivia Wiles +3
Recent work argues that robust training requires substantially larger datasets than those required for standard classification. On CIFAR-10 and CIFAR-100, this translates into a si…
AutoNovel: Automatically Discovering and Learning Novel Visual Categories
Kai Han, Sylvestre-Alvise Rebuffi, Sébastien Ehrhardt +2
We tackle the problem of discovering novel classes in an image collection given labelled examples of other classes. We present a new approach called AutoNovel to address this probl…
Defending Against Image Corruptions Through Adversarial Augmentations
Dan A. Calian, Florian Stimberg, Olivia Wiles +4
Modern neural networks excel at image classification, yet they remain vulnerable to common image corruptions such as blur, speckle noise or fog. Recent methods that focus on this p…
Fixing Data Augmentation to Improve Adversarial Robustness
Sylvestre-Alvise Rebuffi, Sven Gowal, Dan A. Calian +3
Adversarial training suffers from robust overfitting, a phenomenon where the robust test accuracy starts to decrease during training. In this paper, we focus on both heuristics-dri…
LSD-C: Linearly Separable Deep Clusters
Sylvestre-Alvise Rebuffi, Sebastien Ehrhardt, Kai Han +2
We present LSD-C, a novel method to identify clusters in an unlabeled dataset. Our algorithm first establishes pairwise connections in the feature space between the samples of the…
There and Back Again: Revisiting Backpropagation Saliency Methods
Sylvestre-Alvise Rebuffi, Ruth Fong, Xu Ji +1
Saliency methods seek to explain the predictions of a model by producing an importance map across each input sample. A popular class of such methods is based on backpropagating a s…
Automatically Discovering and Learning New Visual Categories with Ranking Statistics
Kai Han, Sylvestre-Alvise Rebuffi, Sebastien Ehrhardt +2
We tackle the problem of discovering novel classes in an image collection given labelled examples of other classes. This setting is similar to semi-supervised learning, but signifi…
NormGrad: Finding the Pixels that Matter for Training
Sylvestre-Alvise Rebuffi, Ruth Fong, Xu Ji +2
The different families of saliency methods, either based on contrastive signals, closed-form formulas mixing gradients with activations or on perturbation masks, all focus on which…
Semi-Supervised Learning with Scarce Annotations
Sylvestre-Alvise Rebuffi, Sebastien Ehrhardt, Kai Han +2
While semi-supervised learning (SSL) algorithms provide an efficient way to make use of both labelled and unlabelled data, they generally struggle when the number of annotated samp…
Efficient parametrization of multi-domain deep neural networks
Sylvestre-Alvise Rebuffi, Hakan Bilen, Andrea Vedaldi
A practical limitation of deep neural networks is their high degree of specialization to a single task and visual domain. Recently, inspired by the successes of transfer learning,…