Explainable deepfake detection extends binary classification by requiring models to not only predict authenticity but also provide interpretable justifications. This expanded scope is critical in practice, where users like forensic analysts need insight into the rationale behind the detection. Despite advancements, current approaches suffer from two critical deficiencies: (1)vulnerability to image quality degradation: detection accuracy plummets on low-quality samples, while naive augmentation strategies may induce feature drift and impair performance as diversity expands. (2) factually flawed explanations: explanation models may omit manipulation evidence or hallucinate irrelevant details, undermining interpretability. To address it, we propose a framework with two innovations. For robust deepfake detection, we introduce Feature-robust Augmentation, which comprises diversified degradation-aware augmentation strategies, and a supervised contrastive learning pattern paired with a mean-teacher architecture that stabilizes features against augmentations through consistency constraints. For explanation, we devise an evidence-grounded preference optimization process that guides model to prioritize genuine manipulation traces by learning from chosen-rejected explanation pairs, where rejected samples are constructed via evidence omission or irrelevant information injection. The proposed approach wins the first place in ACM Multimedia 2026 Explainable Deepfake Detection Challenge.The code is available at https://github.com/oceanflowlab/EDD.git.
Figures & tables
Figure 1. Illustration of analysis for detection robustness. (a) Performance drops under quality degradation. (b) Mean-teacher stabilizes learning under diverse augmentations while the baseline suffers from feature drift.
Figure 2. Overview of our proposed framework. Given an input image, the visual backbone extracts features and fed into a classifier for prediction. During training, the detection branch incorporates degradation-aware augmentations, supervised contrastive learning, and mean-teacher consistency regularization to enhance robustness against quality degradation and feature drift. For explanation generation, the image and deepfake label are sent to the MLLM for generation. To enhance evidence accuracy and completeness, we propose an evidence-grounded optimization process, which generates evidence-omitted and irrelevant information induced explanation to construct rejected samples, guiding the model via the DPO algorithm. We also provide a concise explanation model that is trained via GRPO with tailored semantic fidelity and conciseness rewards.
3 types: per-channel, grayscale, correlated ; two regimes: σ∼U(2,100) and σ∼U(80,100)
Non-Gaussian Noise
Speckle ( σ∼U(2,25) )/ Poisson
JPEG Compression
Quality factor ∼U(10,95)
Color Enhance
Brightness or contrast factor ∼U(0.5,1.5)
Table 1. Summary of degradation strategies and intensities.
Team name
Overall Score
Detection F1
Complex BERT
Simple Overall
Explanation Score
Rank
Pixel Slueth(Ours)
0.841617
0.942399
0.706168
0.775501
0.740834
1
Team Antvengers
0.838154
0.947941
0.696179
0.760555
0.728367
2
MSUteam
0.8312
0.933982
0.70113
0.755707
0.728419
3
HIT VIRLAB
0.822242
0.926326
0.698713
0.737602
0.718157
4
Team1
0.814503
0.916683
0.696807
0.727839
0.712323
5
Qwen3-VL-8B ( Narang et al., 2026b )
0.642776
0.634234
0.664774
0.637863
0.651318
-
Table 2. Public leaderboard Results of Top-5 Teams. Qwen3-VL-8B/InternVL3.5-13B are challenge ( Narang et al., 2026b ) baselines.
Team
Detection F1
Complex Bert
Simple Bert
SLE Score
Entity F1
Claim F1
Explain Score
Overall Score
Rank
Pixel Sleuth(Ours)
0.9424
0.7063
0.6819
0.9782
0.5280
0.4205
0.5800
0.7612
1
Antvengers
0.9479
0.6958
0.6699
0.9720
0.5078
0.3940
0.5618
0.7549
2
MSUteam
0.9340
0.7004
0.6509
0.9726
0.4987
0.3932
0.5571
0.7456
3
Team1
0.9167
0.6959
0.6321
0.9335
0.5155
0.4021
0.5590
0.7378
4
HIT VIRLAB
0.9263
0.6988
0.6493
0.9287
0.4467
0.3436
0.5235
0.7249
5
Table 3. Final hidden-test results for Top-5 team.
Figure 3. Visualizations for deepfake explanation. Result of baseline ( Yang et al., 2025 ) shows irrelevant information, fails to identify valid evidence. SFT-only model captures partial evidence (marked in green) but still suffers irrelevant details(marked in red). Our model generates complete and accurate explanation(marked by green) without introducing spurious information.
Figure 4. Qualitative comparison of detection robustness under samples with various image degradations. The baseline model misclassifies degraded samples, while our model shows correct predictions across different degradation types.
Configuration
Detection F1
Detection Accuracy
Baseline
0.8824
0.8877
+ DA
0.9402
0.9411
+ DA& SC
0.9588
0.9623
+ DA& MT
0.9593
0.9627
Full(+ DA& MT & SC)
0.9724
0.9793
Table 4. Ablation on detection components.
Training Strategy
Fact F1
Claim F1
BERTScore
Baseline
0.4402
0.3362
0.6038
+ SFT
0.5173
0.4218
0.6673
+ DPO (Omit Only)
0.5666
0.4860
0.6912
+ DPO (Irrelevant Only)
0.5743
0.5042
0.6856
+ DPO (Irrelevant + Omit )
0.5942
0.5144
0.7131
Full(DPO with all three types)
0.5969
0.5227
0.7184
Table 5. Ablation on complex explanation model designs.