ARIA: Adversarially Robust Image Attribution for Content Provenance

Andriushchenko, Maksym; Li, Xiaoyang Rebecca; Oxholm, Geoffrey; Gittings, Thomas; Bui, Tien D.; Flammarion, Nicolas; Collomosse, John

doi:10.1109/cvprw56347.2022.00013

Cited by 2 publications

(2 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Due to the negative sign, the criterion prefers pseudo-samples that lead to flatter maxima of the likelihood. In line with recent insights into sharp and flat minima of loss surfaces [Dinh et al, 2017, Li et al, 2018, Andriushchenko and Flammarion, 2022, such a penalty can be expected to improve generalization. The lower the curvature, the more probability mass (area under the likelihood) is expected on…”

Section: Approximate Selection Criteriamentioning

confidence: 64%

Approximate Bayes Optimal Pseudo-Label Selection

Rodemann¹,

Goschenhofer²,

Dorigatti³

et al. 2023

Preprint

View full text Add to dashboard Cite

Semi-supervised learning by self-training heavily relies on pseudo-label selection (PLS). The selection often depends on the initial model fit on labeled data. Early overfitting might thus be propagated to the final model by selecting instances with overconfident but erroneous predictions, often referred to as confirmation bias. This paper introduces BPLS, a Bayesian framework for PLS that aims to mitigate this issue. At its core lies a criterion for selecting instances to label: an analytical approximation of the posterior predictive of pseudo-samples. We derive this selection criterion by proving Bayes optimality of the posterior predictive of pseudo-samples. We further overcome computational hurdles by approximating the criterion analytically. Its relation to the marginal likelihood allows us to come up with an approximation based on Laplace's method and the Gaussian integral. We empirically assess BPLS for parametric generalized linear and non-parametric generalized additive models on simulated and realworld data. When faced with high-dimensional data prone to overfitting, BPLS outperforms traditional PLS methods. a a Code available at: https://anonymous.

show abstract