Boosting Robust AIGI Detection with LoRA-based Pairwise Training

TL;DR

This paper introduces a LoRA-based pairwise training method to enhance the robustness of AI-generated image detection under severe distortions, outperforming existing methods in wild scenarios.

Contribution

The paper proposes a novel finetuning and pairwise training strategy using a visual foundation model to improve AIGI detection robustness against complex distortions.

Findings

Achieved top 3 placement in NTIRE Robust AI-Generated Image Detection challenge.

Enhanced detection performance on distorted images compared to baseline models.

Utilized distortion and size simulations to better match real-world data distribution.

Abstract

The proliferation of highly realistic AI-Generated Image (AIGI) has necessitated the development of practical detection methods. While current AIGI detectors perform admirably on clean datasets, their detection performance frequently decreases when deployed "in the wild", where images are subjected to unpredictable, complex distortions. To resolve the critical vulnerability, we propose a novel LoRA-based Pairwise Training (LPT) strategy designed specifically to achieve robust detection for AIGI under severe distortions. The core of our strategy involves the targeted finetuning of a visual foundation model, the deliberate simulation of data distribution during the training phase, and a unique pairwise training process. Specifically, we introduce distortion and size simulations to better fit the distribution from the validation and test sets. Based on the strong visual representation capability of the visual foundation model, we finetune the model to achieve AIGI detection. The pairwise training is utilized to improve the detection via decoupling the generalization and robustness optimization. Experiments show that our approach secured the 3th placement in the NTIRE Robust AI-Generated Image Detection in the Wild challenge