Paper Image

Vision-language models detect fake images

Published on:

26 October 2023

Primary Category:

Computer Vision and Pattern Recognition

Paper Authors:

You-Ming Chang,

Chen Yeh,

Wei-Chen Chiu,

Ning Yu

Bullets

Key Details

Proposes using vision-language models for deepfake detection

Formulates it as a visual question answering problem

Employs prompt tuning to optimize models for the task

Achieves high accuracy on held-out datasets

AI generated summary

Vision-language models detect fake images

This paper proposes using vision-language models to detect fake images generated by deep generative models. They formulate it as a visual question answering task, asking 'Is this photo real?', and use prompt tuning techniques to optimize the model. Experiments show high accuracy on held-out datasets, demonstrating the approach's effectiveness and transferability.

Answers from this paper

Comments

No comments yet, be the first to start the conversation...

Sign up to comment on this paper

Sign Up