Evolution of Detection Performance throughout the Online Lifespan of Synthetic Images - Centre Borelli UMR 9010
Communication Dans Un Congrès Année : 2024

Evolution of Detection Performance throughout the Online Lifespan of Synthetic Images

Résumé

Synthetic images disseminated online significantly differ from those used during the training and evaluation of the state-of-the-art detectors. In this work, we analyze the performance of synthetic image detectors as deceptive synthetic images evolve throughout their online lifespan. Our study reveals that, despite advancements in the field, current state-of-the-art detectors struggle to distinguish between synthetic and real images in the wild. Moreover, we show that the time elapsed since the initial online appearance of a synthetic image negatively affects the performance of most detectors. Ultimately, by employing a retrieval-assisted detection approach, we demonstrate the feasibility to maintain initial detection performance throughout the whole online lifespan of an image and enhance the average detection efficacy across several state-of-the-art detectors by 6.7% and 7.8% for balanced accuracy and AUC metrics, respectively.
Fichier principal
Vignette du fichier
ECCVW24___DMIWinterSchool24_Synthetic_Image_Detection__HAL_-1.pdf (3.66 Mo) Télécharger le fichier

Dates et versions

hal-04735891 , version 1 (14-10-2024)

Identifiants

Citer

Dimitrios Karageorgiou, Quentin Bammey, Valentin Porcellini, Bertrand Goupil, Denis Teyssou, et al.. Evolution of Detection Performance throughout the Online Lifespan of Synthetic Images. European Conference on Computer Vision, IEEE / CVF, Sep 2024, Milan (Italie), Italy. ⟨10.48550/arXiv.2408.11541⟩. ⟨hal-04735891⟩
13 Consultations
11 Téléchargements

Altmetric

Partager

More