Rethinking Cross-Domain Evaluation for Face Forgery Detection with Semantic Fine-grained Alignment and Mixture-of-Experts

• URL: http://arxiv.org/abs/2604.21478v1
• Date: Thu, 23 Apr 2026 09:39:33 GMT
• Title: Rethinking Cross-Domain Evaluation for Face Forgery Detection with Semantic Fine-grained Alignment and Mixture-of-Experts
• Abstract summary: Existing face forgery detectors can't achieve satisfactory performance because of poor generalization ability across datasets.<n>We propose textbfCross-AUC, an evaluation metric that can compute AUC across dataset pairs.<n>We also propose the novel framework textbfSemantic textbfFine-grained textbfAlignment and textbfMixture-of-Experts.
• Score: 22.1146135455996
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Nowadays, visual data forgery detection plays an increasingly important role in social and economic security with the rapid development of generative models. Existing face forgery detectors still can't achieve satisfactory performance because of poor generalization ability across datasets. The key factor that led to this phenomenon is the lack of suitable metrics: the commonly used cross-dataset AUC metric fails to reveal an important issue where detection scores may shift significantly across data domains. To explicitly evaluate cross-domain score comparability, we propose \textbf{Cross-AUC}, an evaluation metric that can compute AUC across dataset pairs by contrasting real samples from one dataset with fake samples from another (and vice versa). It is interesting to find that evaluating representative detectors under the Cross-AUC metric reveals substantial performance drops, exposing an overlooked robustness problem. Besides, we also propose the novel framework \textbf{S}emantic \textbf{F}ine-grained \textbf{A}lignment and \textbf{M}ixture-of-Experts (\textbf{SFAM}), consisting of a patch-level image-text alignment module that enhances CLIP's sensitivity to manipulation artifacts, and the facial region mixture-of-experts module, which routes features from different facial regions to specialized experts for region-aware forgery analysis. Extensive qualitative and quantitative experiments on the public datasets prove that the proposed method achieves superior performance compared with the state-of-the-art methods with various suitable metrics.

Related papers

• When AUC Misleads: Polarization-Aware Evaluation of Deepfake Detectors under Domain Shift [20.13000766160115]
  Deepfake detection has become an active research area, with recent methods focusing on improving generalization to unseen manipulations.<n>We introduce a novel metric, Cross-dataset AUC (Cross-AUC), that averages per-domain AUCs with a measure of prediction polarization.<n>Cross-AUC not only assesses the generalization capabilities of deepfake detectors under domain shifts more realistically, but it is also interpretable as it better explains the reason behind a drop in performance.
  arXiv  Detail & Related papers  (2026-06-17T15:24:06Z)
• K-Models: a Flexible and Interpretable Method for Ordinal Clustering with Application to Antigen-Antibody Interaction Profiles [65.21921642886367]
  K-Models is a novel framework that integrates ordinal constraints and estimates key underlying elements of the random process generating the observed functional profiles.<n>The proposed method is evaluated through simulations and real-world applications.
  arXiv  Detail & Related papers  (2026-05-14T13:35:44Z)
• ConformalSAM: Unlocking the Potential of Foundational Segmentation Models in Semi-Supervised Semantic Segmentation with Conformal Prediction [57.930531826380836]
  This work explores whether a foundational segmentation model can address label scarcity in the pixel-level vision task as an annotator for unlabeled images.<n>We propose ConformalSAM, a novel SSSS framework which first calibrates the foundation model using the target domain's labeled data and then filters out unreliable pixel labels of unlabeled data.
  arXiv  Detail & Related papers  (2025-07-21T17:02:57Z)
• Cross-Domain Learning for Video Anomaly Detection with Limited Supervision [12.290352736331602]
  Video Anomaly Detection (VAD) automates the identification of unusual events, such as security threats in surveillance videos. Existing cross-domain VAD methods focus on unsupervised learning, resulting in performance that falls short of real-world expectations. We introduce a novel weakly-supervised framework for Cross-Domain Learning (CDL) in VAD that incorporates external data during training by estimating its prediction bias and adaptively minimizing that using the predicted uncertainty.
  arXiv  Detail & Related papers  (2024-08-09T17:28:29Z)
• Downstream-Pretext Domain Knowledge Traceback for Active Learning [138.02530777915362]
  We propose a downstream-pretext domain knowledge traceback (DOKT) method that traces the data interactions of downstream knowledge and pre-training guidance. DOKT consists of a traceback diversity indicator and a domain-based uncertainty estimator. Experiments conducted on ten datasets show that our model outperforms other state-of-the-art methods.
  arXiv  Detail & Related papers  (2024-07-20T01:34:13Z)
• Masked Gamma-SSL: Learning Uncertainty Estimation via Masked Image Modeling [19.000718685399935]
  This work proposes a semantic segmentation network that produces high-quality uncertainty estimates in a single forward pass. We exploit general representations from foundation models and unlabelled datasets through a Masked Image Modeling (MIM) approach. For neural networks used in safety-critical applications, bias in the training data can lead to errors.
  arXiv  Detail & Related papers  (2024-02-27T15:49:54Z)
• CrossDF: Improving Cross-Domain Deepfake Detection with Deep Information Decomposition [53.860796916196634]
  We propose a Deep Information Decomposition (DID) framework to enhance the performance of Cross-dataset Deepfake Detection (CrossDF) Unlike most existing deepfake detection methods, our framework prioritizes high-level semantic features over specific visual artifacts. It adaptively decomposes facial features into deepfake-related and irrelevant information, only using the intrinsic deepfake-related information for real/fake discrimination.
  arXiv  Detail & Related papers  (2023-09-30T12:30:25Z)
• Towards General Visual-Linguistic Face Forgery Detection [95.73987327101143]
  Deepfakes are realistic face manipulations that can pose serious threats to security, privacy, and trust. Existing methods mostly treat this task as binary classification, which uses digital labels or mask signals to train the detection model. We propose a novel paradigm named Visual-Linguistic Face Forgery Detection(VLFFD), which uses fine-grained sentence-level prompts as the annotation.
  arXiv  Detail & Related papers  (2023-07-31T10:22:33Z)
• Self-Supervised Graph Transformer for Deepfake Detection [1.8133635752982105]
  Deepfake detection methods have shown promising results in recognizing forgeries within a given dataset. Deepfake detection system must remain impartial to forgery types, appearance, and quality for guaranteed generalizable detection performance. This study introduces a deepfake detection framework, leveraging a self-supervised pre-training model that delivers exceptional generalization ability.
  arXiv  Detail & Related papers  (2023-07-27T17:22:41Z)
• Cluster-level pseudo-labelling for source-free cross-domain facial expression recognition [94.56304526014875]
  We propose the first Source-Free Unsupervised Domain Adaptation (SFUDA) method for Facial Expression Recognition (FER) Our method exploits self-supervised pretraining to learn good feature representations from the target data. We validate the effectiveness of our method in four adaptation setups, proving that it consistently outperforms existing SFUDA methods when applied to FER.
  arXiv  Detail & Related papers  (2022-10-11T08:24:50Z)
• MD-CSDNetwork: Multi-Domain Cross Stitched Network for Deepfake Detection [80.83725644958633]
  Current deepfake generation methods leave discriminative artifacts in the frequency spectrum of fake images and videos. We present a novel approach, termed as MD-CSDNetwork, for combining the features in the spatial and frequency domains to mine a shared discriminative representation.
  arXiv  Detail & Related papers  (2021-09-15T14:11:53Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.