ODDN: Addressing Unpaired Data Challenges in Open-World Deepfake Detection on Online Social Networks

TL;DR

This paper introduces ODDN, a novel deepfake detection network designed to effectively handle unpaired and compressed data in open-world social media scenarios, improving robustness and detection accuracy.

Contribution

The paper proposes ODDN with two modules, ODA and CGC, to address unpaired data challenges and compression variability in deepfake detection, a problem not well handled by existing methods.

Findings

ODDN outperforms state-of-the-art methods on 17 datasets.

Effective aggregation of unpaired data improves detection robustness.

Gradient correction enhances performance across diverse compression methods.

Abstract

Despite significant advances in deepfake detection, handling varying image quality, especially due to different compressions on online social networks (OSNs), remains challenging. Current methods succeed by leveraging correlations between paired images, whether raw or compressed. However, in open-world scenarios, paired data is scarce, with compressed images readily available but corresponding raw versions difficult to obtain. This imbalance, where unpaired data vastly outnumbers paired data, often leads to reduced detection performance, as existing methods struggle without corresponding raw images. To overcome this issue, we propose a novel approach named the open-world deepfake detection network (ODDN), which comprises two core modules: open-world data aggregation (ODA) and compression-discard gradient correction (CGC). ODA effectively aggregates correlations between compressed and raw samples through both fine-grained and coarse-grained analyses for paired and unpaired data, respectively. CGC incorporates a compression-discard gradient correction to further enhance performance across diverse compression methods in OSN. This technique optimizes the training gradient to ensure the model remains insensitive to compression variations. Extensive experiments conducted on 17 popular deepfake datasets demonstrate the superiority of the ODDN over SOTA baselines.