Not Every Cue Is Evidence: Cue Supportability for Forensic Distillation in Face Forgery Detection

Binjia Zhou; Dawei Luo; Shuai Chen; Feng Xu; Haoyuan Li; Jia wen Seow; Jiachi Wang; Jie Lei

cs.CVOct 12, 2026

Not Every Cue Is Evidence: Cue Supportability for Forensic Distillation in Face Forgery Detection

Authors: Binjia Zhou, Dawei Luo, Shuai Chen, Feng Xu, Haoyuan Li, Jia wen Seow, Jiachi Wang, Jie Lei, +2 more

Organizations: School of Software Technology, Zhejiang University · Ant Group · College of Computer Science and Technology, Zhejiang University · College of Computer Science and Technology, Zhejiang University of Technology

Abstract

Explainable face forgery detection requires not only correct judgments but also forensic evidence supported by the input image. However, training cues are typically derived from the forgery generation pipeline. Since images available to the model are of limited quality and resolution, it is difficult to observe the fine-grained traces these cues refer to. Training methods that further supply such cues as privileged supervision exacerbate the problem, as the student is taught to articulate evidence it cannot perceive. The student consequently resorts to templated or over-generalized expressions, giving rise to the correct-answer, wrong-evidence (CAWE) failure. We propose a dual supportability audit that filters privileged cues before distillation. Each candidate cue must be predictable from the remaining privileged context and causally tied to high-attribution visual regions in the current image. Only cues satisfying both conditions receive substantial distillation weight, preventing unsupported details from leaking into the student. Experiments across eight benchmarks show that our method maintains or improves detection accuracy while significantly raising explanation faithfulness compared to existing paradigms.

Figures & tables

Figure 1: Comparison of Answer Accuracy, Cue Precision, and CAWE on FF++ across training paradigms and our ablation variants.

Figure 2: Comparison between conventional FFD supervision and our method. Conventional approaches train models with privileged cues that cannot be observed from the image, leading to template-like evidence. Our supportability audit filters unsupported cues, enabling image-specific faithful explanations.

Figure 3: Overview of our framework. The transfer utility (TU) module (top) masks each candidate cue and measures the likelihood drop, down-weighting cues trivially copied from the annotation. The visual sensitivity (VS) module (bottom) perturbs high-attribution image regions and checks whether cue likelihood is sensitive to visual content. Cues that fail either check receive near-zero weight in the distillation loss.

Figure 4: CAWE on FF++ across training paradigms; higher means more unfaithful explanations. Most existing methods see CAWE rise with training, while ours consistently reduces it.

		FF++				Celeb-DF v2				DFDC				WildDF
Category	Method	ACC	AUC	EER $\downarrow$	CAWE $\downarrow$	ACC	AUC	EER $\downarrow$	CAWE $\downarrow$	ACC	AUC	EER $\downarrow$	CAWE $\downarrow$	ACC	AUC	EER $\downarrow$	CAWE $\downarrow$
Traditional	Xception	95.73	96.30	9.52	–	90.34	89.75	15.20	–	79.32	81.58	24.50	–	75.26	80.89	26.80	–
	SBI	97.80	99.64	2.40	–	92.45	93.18	11.60	–	82.50	87.30	19.80	–	78.40	84.50	22.50	–
	Forensics Adapter	98.35	99.71	1.82	–	94.20	95.80	9.40	–	85.10	90.20	16.50	–	80.60	87.30	19.80	–
Explainable	FFAA	94.80	96.50	8.75	62.8	88.50	90.20	16.80	68.4	76.30	80.50	26.20	65.2	72.40	78.60	28.50	71.6
	VLFFD	97.20	99.23	3.85	56.3	93.80	95.40	10.20	60.1	83.60	88.50	18.40	62.4	79.50	85.80	21.30	65.8

Table 1: Main results on four benchmarks. Color intensity indicates SOTA ranking. The Gain row reports improvements over the strongest baseline (reductions for EER and CAWE); a negative value indicates underperformance. CAWE is reported only for methods that generate forensic evidence. Our method achieves competitive detection accuracy while substantially reducing CAWE over all explainable baselines.

Method	Tok. $\downarrow$	Inf. $\downarrow$	ACC	CAWE $\downarrow$	Tok. $\downarrow$	Inf. $\downarrow$	ACC	CAWE $\downarrow$	Tok. $\downarrow$	Inf. $\downarrow$	ACC	CAWE $\downarrow$	Tok. $\downarrow$	Inf. $\downarrow$	ACC	CAWE $\downarrow$
	DeeperForensics				Celeb-DF v2				ForgeryNet				Celeb-DF++
Standard SFT	6.8M	842	89.45	74.2	4.2M	856	91.20	71.6	12.4M	810	74.80	79.8	5.6M	838	90.65	73.5
Self-Correction SFT	15.2M	1,104	90.80	76.5	9.4M	1,126	92.60	73.4	27.8M	1,068	76.20	81.4	12.5M	1,092	91.85	75.8
LLaVA-CoT	32.1M	1,158	92.30	66.8	19.8M	1,182	94.80	64.2	58.6M	1,120	80.40	74.5	26.4M	1,146	94.10	67.6
Standard GRPO	89.4M	1,318	93.80	70.4	55.2M	1,346	95.60	68.2	163.2M	1,274	82.40	75.8	73.6M	1,302	95.20	72.6
DAPO	114.7M	1,280	94.10	71.8	70.8M	1,304	95.30	69.4	209.4M	1,238	81.90	78.2	94.4M	1,266	94.80	73.0

Table 2: Comparison across paradigms on four benchmarks. Tok. denotes student-side training tokens (rollouts and student forward passes) drawn from each benchmark’s training split; teacher and audit passes, which roughly double our per-step cost, are excluded. Inf. denotes average inference tokens per sample. Ours achieves the highest accuracy and lowest CAWE across datasets with fewer student-side tokens than all other distillation and RL alternatives and shorter responses. Shading follows Table 1 ; fewer tokens are better.

Figure 5: Qualitative comparison on two forged samples; purple text lists the annotated artifacts. Existing methods either predict the wrong answer or produce template-like evidence unrelated to the specific image. Our method generates precise, image-specific forensic cues that are visually grounded and support the decision.

Figure 6: Human-judged support ratio (%) of forensic cues on six benchmarks (150 image-cue pairs per cue source and benchmark). Error bars: 95% bootstrap CIs.

Figure 7: Ablation results on FF++, Celeb-DF v2, DFDC, and overall average. (a) Core component ablation. (b) Training paradigm comparison. (c) Distillation target variants. (d) Perturbation operator choices. (e) Attribution top- $k$ sensitivity. (f) Answer accuracy under component removal.

Percentile $w_{c}$	# Cues	VLM Sup. $\uparrow$	Human Sup. $\uparrow$
0 – 15%	150	10.0	14.7
15 – 30%	150	24.0	28.7
30 – 45%	150	36.7	42.0
45 – 60%	150	50.7	54.7
60 – 75%	150	64.0	68.0
75 – 90%	150	76.7	78.7

Table 3: Supportability score validation on FF++ (1,000 cues). Higher

w_{c}

bins yield monotonically higher VLM and human support rates (%).

Appendix figures & tables13 assets

Supplementary material from the paper’s appendix.

Appendix

Information	Teacher	Student
Image $I$	✓	✓
Task prompt $q$	✓	✓
GT label $y$	✓	✗
Forgery type $o$	✓	✗
Manipulation region $r$	✓	✗
Target cue $c$	Base: ✓ / Masked: ✗	✗

Appendix

Table 4: Information accessibility for teacher and student.

	Acc	Prec.	Rec.	F1	$\kappa$
CAWE Eval.	87.4	91.0	89.5	90.2	0.72

Appendix

Table 5: Reliability of the CAWE evaluator against human majority vote on 3,600 image-cue pairs. Metrics (%) are macro-averaged over six benchmarks;

\kappa

is Cohen’s kappa.

Method	FF++		Celeb-DF v2		DFDC		WildDF		Average
	Orig.	1-Cue	Orig.	1-Cue	Orig.	1-Cue	Orig.	1-Cue	Orig.	1-Cue
FFAA	62.8	49.8	68.4	54.2	65.2	51.7	71.6	57.4	67.0	53.3
VLFFD	56.3	44.7	60.1	48.2	62.4	49.5	65.8	52.6	61.2	48.8
Identity-Aware VLM	53.6	41.8	55.4	43.1	61.7	48.7	59.2	46.0	57.5	44.9
RLSBI	58.1	45.9	63.8	50.7	62.0	49.3	68.4	54.6	63.1	50.1
SIDA	54.5	43.0	58.6	46.8	63.8	50.2	61.0	48.0	59.5	47.0

Appendix

Table 6: CAWE (%) under the original evaluation and the one-cue control. “1-Cue” retains one uniformly sampled cue per correctly classified sample and averages results over five random seeds.

Setting	Value
Optimizer	AdamW
Weight decay	0.01
Scheduler	Cosine, 3% warmup
Grad. clipping	Max norm 1.0
Precision	BF16
Max seq. length	2048 tokens

Appendix

Table 7: Implementation summary. Main-text hyperparameters (

\tau_{a}

m_{\mathrm{pert}}

T_{\mathrm{pert}}

k

\lambda_{\mathrm{ans}}

\lambda_{\mathrm{sup}}

, epochs, LR, batch size, hardware) are omitted.

Dataset	Resolution	Vis. Sup. (%)	#Sampled
FF++ Rossler et al. (2019)	$256{\times}256$	51.6	200
Celeb-DF v2 Li et al. (2020b)	$256{\times}256$	40.3	160
DFDC Dolhansky et al. (2020)	$320{\times}320$	30.5	240
WildDF Zi et al. (2020)	varies	21.4	180
DeeperForensics Jiang et al. (2020)	$256{\times}256$	35.2	200
ForgeryNet He et al. (2021)	$256{\times}256$	25.7	220

Appendix

Table 8: Per-dataset image resolution and visual-support ratio. “Vis. Sup.” denotes the percentage, over all annotated cues on the sampled images, of cues whose visual evidence is identifiable by human inspection. #Sampled is the number of randomly sampled test images inspected.

Qwen2.5-VL family Bai et al. (2025b)
Method	Teacher	Params	ACC	CAWE
Standard OPD	Qwen2.5-VL-3B	3B	95.40	61.4
Standard OPD	Qwen2.5-VL-7B	7B	96.20	59.3
Qwen3-VL family Bai et al. (2025a)
Standard OPD	Qwen3-VL-8B	8B	96.50	58.6
Standard OPD	Qwen3-VL-32B	32B	97.30	54.9

Appendix

Table 9: Effect of teacher identity and scale on FF++ (student: Qwen3-VL-8B; same protocol as Table 1 ).

Figure 8: Visualization on a real image (Case 1). $\times$ : wrong prediction; $\checkmark\!\times$ : correct prediction with unfaithful evidence; $\checkmark$ : correct prediction with faithful evidence.

Figure 9: Visualization on a forged image (Case 2). Symbols follow Fig. 8 .

Figure 10: Before vs. after supportability-aware training. Left: explanations before training; right: explanations after training. $\times$ : wrong prediction; $\checkmark\!\times$ : correct prediction with unfaithful evidence; $\checkmark$ : correct prediction with faithful evidence.

		DeeperForensics		Celeb-DF v2		ForgeryNet		Celeb-DF++		DF40
Category	Method	ACC	AUC	ACC	AUC	ACC	AUC	ACC	AUC	ACC	AUC
Traditional	EfficientNet Tan and Le (2019)	88.60	92.40	90.80	93.20	70.40	76.80	89.60	91.80	72.50	78.40
	Face X-ray Li et al. (2020a)	90.40	93.80	92.60	94.50	72.80	78.60	91.80	93.20	74.20	80.60
	F3Net Qian et al. (2020)	91.20	94.60	93.40	95.80	74.50	80.40	92.60	94.50	75.80	82.40
	MAT Zhao et al. (2021)	91.80	95.10	93.80	96.20	75.20	81.60	93.00	94.80	76.40	83.20
	SPSL Liu et al. (2021)	90.80	94.20	93.20	95.60	73.80	79.80	92.40	94.20	75.20	81.80

Appendix

Table 10: Extended comparison on five benchmarks with methods not included in Table 1 . All models are trained on the combined multi-dataset split, following the same protocol as Table 1 .

Category	Method	CAWE $\downarrow$ (Avg.)	$\Delta$ vs. Ours
Explainable	RAIDx Li et al. (2025)	58.6	+27.4
	So-Fake Huang et al. (2025b)	56.4	+25.2
	REVEAL Praharaj et al. (2025)	52.8	+21.6
	FP-DAPO Zhu et al. ()	48.6	+17.4
	GE-AIGC Guo et al. (2026)	46.8	+15.6
	Ours	31.2	–

Appendix

Table 11: Average CAWE across the five benchmarks for explainable methods.

\Delta

denotes the gap relative to our method.

Cat.	Method	Params	Out Tok.	Latency (s)	Thru. (samples/s)
Traditional	Xception	23M	–	0.008	125.0
	SBI	23M	–	0.012	83.3
	Forensics Adapter	151M	–	0.015	66.7
Explainable	FFAA	8B	286	3.82	0.26
	VLFFD	8B	245	3.28	0.30
	Identity-Aware VLM	8B	312	4.16	0.24

Appendix

Table 12: Inference efficiency comparison on FF++. Params: model parameters. Out Tok.: average output tokens. Latency: seconds per sample. Thru.: throughput (samples/second). Traditional methods do not generate text.

Method	Vis. Enc. (s)	Text Gen. (s)	Total (s)	Gen. Ratio
FFAA	0.42	3.40	3.82	89.0%
VLFFD	0.42	2.86	3.28	87.2%
EDD	0.42	2.56	2.98	85.9%
PATE-Forensics	0.42	3.82	4.24	90.1%
FakeShield	0.48	5.38	5.86	91.8%
Ours	0.42	1.66	2.08	79.8%

Appendix

Table 13: Inference time breakdown. Vis. Enc.: visual encoding time. Text Gen.: autoregressive text generation time. Gen. Ratio: fraction of total time spent on text generation.

Explore similar work

Lightweight Complementary-Cue Fusion for Robust Video Face Forgery Detection
May 27, 2026Sunghwan Baek, Tariq Anwaar, Karanveer Singh +1Frequency-Domain Feature Learning Deepfake Detection
Carnegie Mellon University
Trustworthy Image Authentication using Forensic Knowledge Graphs
Jun 22, 2026Tai D. Nguyen, Matthew C. StammImage Forgery Detection Digital Image Forensics
Drexel University, Philadelphia, PA 19104, USA
EviStress: Evidence Stress Learning for Generalizable Face Forgery Detection with Textual Explanations
Oct 12, 2026Jiuyao Jing, Xuanyu Yi, Yu Zheng +1Image Forgery Detection Deepfake Detection
School of Cyber Engineering, Xidian University, Xi’an, China

May 27, 2026J/K move · Enter open · S save

Lightweight Complementary-Cue Fusion for Robust Video Face Forgery Detection

Sunghwan Baek, Tariq Anwaar, Karanveer Singh, Rita Singh

Carnegie Mellon University

Frequency-Domain Feature Learning Deepfake Detection Multimodal Fusion Feature Fusion

Open paper AskPDF arXiv

Current face video forgery detectors use wide or dual-stream backbones. We show that a single, lightweight fusion of two handcrafted cues can achieve higher accuracy with a much smaller model. Based on the Xception baseline model (21.9 million parameters), we build two detectors: LFWS, which adds a 1x1 convolution to combine a low-frequency Wavelet-Denoised Feature (WDF) with a phase-spectrum channel derived from Spatial-Phase Shallow Learning (SPSL), and LFWL, which merges WDF with Local Binary Patterns (LBP) in the same way. This extra module adds only 292 parameters, keeping the total at 21.9 million, smaller than F3Net (22.5 million) and less than half the size of SRM (55.3 million). Even with this minimal overhead, the fused models increase the average area under the curve (AUC) from 74.8% to 78.6% on FaceForensics++ and from 70.5% to 74.9% on DFDC-Preview, gains of 3.8% and 4.4% over the Xception baseline. They also consistently outperform F3Net, SRM, and SPSL in eight public benchmarks, without extra data or test-time augmentation. These results show that carefully paired, handcrafted features, combined through the lightweight fusion block, can provide competitive robustness at a significantly lower cost than comparable frequency-based detectors. Our findings suggest a need to reevaluate scale-driven design choices in face video forgery detection.

Abstract

Explore similar work

Lightweight Complementary-Cue Fusion for Robust Video Face Forgery Detection

Trustworthy Image Authentication using Forensic Knowledge Graphs

EviStress: Evidence Stress Learning for Generalizable Face Forgery Detection with Textual Explanations