Leveraging Unlabeled Data from Unknown Sources via Dual-Path Guidance for Deepfake Face Detection

TL;DR

This paper introduces DPGNet, a novel deepfake detection approach that leverages unlabeled data from unknown sources by aligning cross-domain features and dynamically generating pseudo-labels, improving generalization in real-world scenarios.

Contribution

The paper proposes a dual-path guided network with text-guided alignment and curriculum-driven pseudo-labeling to utilize unlabeled fake face data from unknown sources effectively.

Findings

DPGNet outperforms existing methods on multiple datasets.

The approach effectively bridges domain gaps between different generative models.

Utilizes unlabeled data to enhance deepfake detection accuracy.

Abstract

Existing deepfake detection methods heavily rely on static labeled datasets. However, with the proliferation of generative models, real-world scenarios are flooded with massive amounts of unlabeled fake face data from unknown sources. This presents a critical dilemma: detectors relying solely on existing data face generalization failure, while manual labeling for this new stream is infeasible due to the high realism of fakes. A more fundamental challenge is that, unlike typical unsupervised learning tasks where categories are clearly defined, real and fake faces share the same semantics, which leads to a decline in the performance of traditional unsupervised strategies. Therefore, there is an urgent need for a new paradigm designed specifically for this scenario to effectively utilize these unlabeled data. Accordingly, this paper proposes a dual-path guided network (DPGNet) to address two key challenges: (1) bridging the domain differences between faces generated by different generative models; and (2) utilizing unlabeled image samples. The method comprises two core modules: text-guided cross-domain alignment, which uses learnable cues to unify visual and textual embeddings into a domain-invariant feature space; and curriculum-driven pseudo-label generation, which dynamically utilizes unlabeled samples. Extensive experiments on multiple mainstream datasets show that DPGNet significantly outperforms existing techniques,, highlighting its effectiveness in addressing the challenges posed by the deepfakes using unlabeled data.