Fooling an Automatic Image Quality Estimator - Archive ouverte HAL Access content directly
Conference Papers Year : 2020

Fooling an Automatic Image Quality Estimator


This paper presents our work on the 2020 MediaEval task: “Pixel Privacy: Quality Camouflage for Social Images". Blind Image Quality Assessment (BIQA) is an algorithm predicting a quality score for any given image. Our task is to modify an image to decrease its BIQA score while maintaining a good perceived quality. Since BIQA is a deep neural network, we worked on an adversarial attack approach of the problem.
Fichier principal
Vignette du fichier
Linkmedia_pprivacy_f.pdf (732.79 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03132891 , version 1 (05-02-2021)


  • HAL Id : hal-03132891 , version 1


Benoit Bonnet, Teddy Furon, Patrick Bas. Fooling an Automatic Image Quality Estimator. MediaEval 2020 - MediaEval Benchmarking Intiative for Multimedia Evaluation, Dec 2020, Online, United States. pp.1-4. ⟨hal-03132891⟩
69 View
92 Download


Gmail Facebook X LinkedIn More