Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

Multimodal Fake News Detection: MFND Dataset and Shallow-Deep Multitask Learning

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Multitask fake-news learner beats prior detectors on four benchmarks

desk verdict MFND dataset is a genuinely useful contribution, but the SOTA claim is refuted by the paper's own tables; worth refereeing for the dataset, conditional on fixes. read the letter →

arxiv 2505.06796 v1 pith:SC6JPWRY submitted 2025-05-11 cs.CV

Ye Zhu , Yunan Wang , Zitong Yu This is my paper · ORCID
classification cs.CV
keywords multimodalfakenewsdetectiondeepfakeimagemanipulationlocalizationtextforgerycontrastivelearningcross-modalfusionmultitaskbenchmarkdataset
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper is trying to establish that the next generation of fake news—images altered by deepfake face generators and text rewritten by multimodal language models—needs a richer benchmark and a richer model than the field's usual binary real/fake label. To make that case, it contributes MFND, a 125,000-sample news dataset built from VisualNews that covers 11 manipulation types and carries four kinds of labels: news-level authenticity, image forgery, text forgery, and a bounding box around the manipulated image region. It then proposes SDML, a shallow-deep multitask learner that first aligns image and text embeddings with a softened contrastive loss, fuses them adaptively, and then runs two branches that recombine unimodal detail with the fused features to predict all four outputs at once. The authors report that SDML outperforms the compared multi-task models on MFND and DGM4 and the compared single-task models on MFND, Weibo, and Twitter. If the dataset is faithful to real online fake news, the paper supplies both a benchmark and an architecture for catching AI-manipulated news rather than merely mismatched image-text pairs.

What carries the argument

The argument is carried by a two-stage inference architecture. In the shallow stage, two pretrained encoders (ViT-B/16 for images, the first eight layers of BERT for text) produce embeddings that are aligned in a shared space by Light Punishment Contrastive Learning (LPCL): a momentum model supplies soft similarity targets so that weakly correlated pairs are not punished as harshly as in one-hot contrastive learning, giving $\mathcal{L}_{LC}=\mathcal{L}_{ITC}+\lambda\mathcal{L}_{MM}$. An Adaptive Cross-Modal Fusion module (ACMF) then computes image-text and text-image cross-attention and blends them with a learnable weight $\alpha$ to form a mutual-modality feature. In the deep stage, a Multi-View Extractor (MVE) concatenates multi-level ViT features and multi-scale CNN features for the image branch, while a Context Aggregator (CA) reuses the last four BERT layers for the text branch; each branch is recombined with the mutual feature and sent to dedicated projections for binary detection, bounding-box regression, and text detection. The total loss is the sum of the contrastive, news binary, image detection/localization, and text detection losses.

What would settle it

A reader could falsify the practical claim by training a simple classifier to separate MFND's generated samples from its real samples using only low-level signals—frequency spectra, blending edges, or text perplexity—and showing it reaches near-perfect accuracy; that would mean the manipulations leave pipeline artifacts SDML could exploit, so its top benchmark scores would not transfer to independently produced fake news. Likewise, holding out an entire manipulation type (an unseen image generator or LLM) and retesting SDML would show whether the model detects manipulation in general or only the specific generators in MFND.

Watch

Extended reading notes

Core claim

The paper's central claim is that multimodal fake news is best handled as a jointly learned set of four tasks—news-level binary authenticity, image forgery detection, image manipulation localization, and text forgery detection—and that a shallow-deep architecture which separates alignment from reasoning achieves the best published numbers on all four at once. On the newly built MFND dataset (125,000 image-text news pairs, 11 manipulation types spanning StyleGAN3, LatTrans, and FaceSwapper image edits and MiNiGPT-v2/CogVLM text rewrites), the proposed SDML model reports 92.43 AUC and 85.54 ACC for news detection, 95.65 AUC for image detection, an IoUmean of 77.83 for grounding the tampered region, and 95.76 AUC for text detection, surpassing three multi-task baselines. The same model also beats previous multi-task methods on DGM4 and previous single-task detectors on MFND, Weibo, and Twitter, which is the evidence for the state-of-the-art claim.

Load-bearing premise

The load-bearing premise is that the automatically generated fakes in MFND—machine-made face edits and LLM text rewrites kept at 50–75% similarity to their images—resemble the fake news a deployed detector would actually meet online, so that the benchmark scores measure real detection skill rather than artifacts of the generation pipeline.

Editorial extensions

If this is right

  • A single model can output a real/fake label for the news item, for the image, and for the text, plus a bounding box for the manipulated image region, and the multitask formulation itself improves accuracy over single-task training.
  • Softening contrastive learning with momentum-derived targets yields better image-text alignment for fake-news detection than standard one-hot contrastive loss.
  • Fake news generated by deepfake face editors and LLM text rewrites is measurably harder than mismatched image-text pairs: all compared methods score lower on MFND than on Weibo and Twitter.
  • The MFND benchmark's four label types and 11 manipulation types provide a reusable evaluation standard for future multimodal detection and grounding work.
  • On the same data, binary news accuracy is higher under the multi-task objective than under single-task training, indicating that the extra supervision stabilizes learned features.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper's claims: if MFND's generation pipeline leaves detectable fingerprints, such as common GAN frequency artifacts or LLM style biases, the reported gains may partly reflect artifact spotting; a cross-generator test that trains on MFND and tests on fake news from unseen generators would show how much of the advantage transfers.
  • The dataset's 50–75% image-text similarity control may make the task artificially consistent within a narrow band; evaluating the same model on deliberately mismatched, highly similar, and perfectly consistent pairs would reveal how sensitive the method is to that design choice.
  • Because the paper deliberately drops text localization and keeps only image bounding boxes, token-level manipulation grounding remains open; SDML's text branch could be extended with a sequence-labeling head to localize altered text spans for explainability.
  • The largest reported gains are in image localization (IoU), which suggests the deep two-branch image path, not the fusion module, drives grounding improvements; isolating that branch could lead to lighter models with similar localization accuracy.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper introduces MFND, a new multimodal fake news detection dataset built from VisualNews image-text pairs and augmented with 11 manipulation types (StyleGAN3, LatTrans, FaceSwapper for images; MiNiGPT-v2 and CogVLM for text), providing binary news labels, image forgery labels, image bounding boxes, and text forgery labels. The authors also propose SDML, a Shallow-Deep Multitask Learning model that combines a light-punishment contrastive alignment with momentum distillation, an adaptive cross-modal fusion module, and a two-branch deep inference stage for joint image/text detection and image localization. The model is evaluated on MFND, DGM4, Weibo, and Twitter under both multi-task and single-task settings, with ablations and visualizations. The central claim is that SDML achieves state-of-the-art detection and localization performance on all four datasets under both settings.

Significance. If the results hold, the work would be a substantial contribution: MFND is larger and more diverse than existing deepfake-oriented multimodal news datasets (125k samples, 11 manipulation types, four news categories), and SDML provides a principled multitask formulation that jointly handles news-level, image-level, and text-level decisions plus grounding. The manuscript describes the architecture with enough detail to reimplement, reports systematic ablations showing each module contributes, and states that code and data are released. The main significance risk is twofold: the headline SOTA claim is contradicted by the paper's own tables on specific metrics, and the realism of the synthetic MFND manipulations is asserted without external validation, which is load-bearing for the dataset contribution.

major comments (3)
  1. [Section 5.2, Table 2 and Table 3] The central claim that SDML achieves state-of-the-art performance on four benchmark datasets under both settings is contradicted by the paper's own numbers. In Table 2 (DGM4, multi-modal multi-task setting), SDML's multimodal ACC is 86.14, lower than the DGM4 baseline's 86.39; in Table 3 (Weibo, single-task setting), SDML's mF1 is 92.44, lower than COOLANT's 92.63. The text in Section 5.2 states that ACC and F1 values are 'significantly better than the other baseline methods on all datasets', which is false as written. The claim must be qualified to specific datasets and metrics, or the tables/code must be corrected.
  2. [Section 3.1] The realism of the MFND dataset is load-bearing for the dataset contribution, but the paper provides no evidence that the generated manipulations resemble real-world fake news. The statement that 'the similarity between image-text sample pairs is controlled to be 50% to 75%' does not specify the similarity metric, and no human study, artifact analysis, or cross-dataset generalization experiment is reported. If the synthetic fakes contain generation-pipeline artifacts, or if the 50-75% similarity control makes the task artificially easy or hard, the measured performance on MFND may not transfer to authentic fake news. The manuscript should add a concrete similarity measure, human evaluation, or a generalization experiment to justify the dataset's validity.
  3. [Section 3.2] The dataset split description is internally inconsistent: the text says 'the 95k pairs are part of model training, another 15k pairs are part of testing and the remaining 15k are for testing', listing 15k twice and omitting a validation split. Please clarify whether the split is train/validation/test and report the exact sizes of each split, as this affects the reproducibility of all reported results.
minor comments (4)
  1. [Section 4.1, Eq. (1) and Eq. (2)] Both equations use the superscript 't→v' even though one defines image-to-text similarity and the other text-to-image similarity; the superscripts should be 'v→t' and 't→v' respectively to match the text.
  2. [Section 4.1, Eq. (4)] The momentum similarity equations also appear to reuse 't→v' for both directions; please correct the notation for the image-to-text direction.
  3. [Table 4] The rows 'w/o Image' and 'w/o Text' report only two metrics (AUC and ACC) whereas all other rows report nine columns; the missing metrics should be filled in or explicitly marked as not applicable.
  4. [Section 5.1] The evaluation metrics list says IoU thresholds 0.5, 0.75, and 0.9 are used to 'calculate the average accuracy', but Table 1 and Table 2 report only IoUmean and IoU50; please clarify how IoU50 relates to the thresholds and whether IoU75/IoU90 results are omitted or averaged somewhere.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: SDML is trained with supervised labels and evaluated on held-out and external benchmarks; internal SOTA metric inconsistencies are correctness concerns, not circular reasoning.

full rationale

I find no circular derivation chain in this paper. The SDML model is trained with supervised labels (binary news label, image manipulation label, text manipulation label, and bounding-box label) under the total loss in Eq. (21) and evaluated on held-out splits of MFND plus external benchmarks DGM4, Weibo, and Twitter. The MFND labels are produced by external generation tools (StyleGAN3, LatTrans, FaceSwapper, MiNiGPT-v2, CogVLM) rather than derived from the model's own outputs, so the detection and localization predictions are not equivalent to the training inputs by construction. The self-citations to prior work by the corresponding author (e.g., FusionMamba, GenFace) appear in related work and are not load-bearing for the claimed result. The internal inconsistency in the SOTA claim—for example, Table 2 shows SDML multimodal ACC 86.14 below DGM4's 86.39 on DGM4, and Table 3 shows Weibo mF1 92.44 below COOLANT's 92.63—is a correctness or reporting concern about the paper's own tables, not circularity. Likewise, the unvalidated realism of automatically generated MFND manipulations is a dataset-validity concern requiring external human or artifact studies, not circular reasoning. No step reduces to its own input by definition or through a self-citation chain.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

The central empirical claim rests on dataset realism and evaluation fairness. The model itself has no invented physical entities. The principal unvalidated inputs are the automatic data generation pipeline and the assumption that baselines were tuned fairly.

free parameters (7)
  • temperature tau (contrastive) = learnable, initialized 0.07
    Scales similarity logits in Eqs. (1)-(4); learned during training and affects the alignment loss.
  • lambda (momentum distillation weight) = learnable, initialized 0.02
    Weights the momentum distillation loss in Eq. (5); set by hand initially.
  • alpha (modal fusion weight) = learnable
    Blends image-text and text-image cross-context features in Eq. (10); adapted during training.
  • momentum queue size M = 65535
    Stores recent image-text pairs for momentum similarity; chosen by hand in Section 5.1.
  • text encoder / context aggregator layer split = 8 encoder layers, 4 aggregator layers
    Selected by ablation in Fig. 4; affects text feature quality.
  • Multi-View Extractor depth/scales = n=m=4
    Number of ViT layers and CNN scales in the multi-view extractor; selected by ablation in Fig. 5.
  • batch size / epochs / optimizer hyperparameters = 64, 100, AdamW lr 5e-6 to 5e-7, weight decay 0.005
    Training configuration in Section 5.1; chosen by hand.
assumptions (6)
  • domain assumption Pretrained ViT-B/16 and BERT provide transferable unimodal features for news images and text.
    Section 4.1 initializes encoders with ImageNet and BERT weights; if these features do not transfer to news manipulation, SDML degrades.
  • domain assumption MFND's GAN/LLM manipulation pipeline produces fake news representative of real-world multimodal manipulations.
    Section 3.1 builds all fake samples with StyleGAN3, LatTrans, FaceSwapper, MiNiGPT-v2, and CogVLM; no human study validates realism.
  • domain assumption The 50-75% image-text similarity constraint for generated text is meaningful and correctly measured.
    Section 3.1 states the constraint but does not specify the similarity metric; loose or incorrect control could make the text forgery task artificially easy or hard.
  • domain assumption Source data from VisualNews filtered to human-centered pairs is a suitable proxy for real news.
    Section 3 selects 200k image-text pairs centered on humans; selection bias may limit generalization.
  • domain assumption Bounding box labels and text forgery labels are accurate as generated.
    Section 3.1 labels fake images with boxes and generated text as fake; no quality control or annotator agreement is reported.
  • domain assumption The compared baselines were configured fairly and equivalently.
    Section 5.2 reports baseline numbers but not their hyperparameter tuning or seeds on MFND; unfair baselines would inflate SDML's relative gains.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Multimodal Fake News Detection: MFND Dataset and Shallow-Deep Multitask Learning." pith.science (2026). https://pith.science/paper/SC6JPWRY

@misc{pith2026250506796,
  author       = {Pith},
  title        = {Pith review of: Multimodal Fake News Detection: MFND Dataset and Shallow-Deep Multitask Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SC6JPWRY}},
  note         = {Machine review of arXiv:2505.06796}
}
read the original abstract

Multimodal news contains a wealth of information and is easily affected by deepfake modeling attacks. To combat the latest image and text generation methods, we present a new Multimodal Fake News Detection dataset (MFND) containing 11 manipulated types, designed to detect and localize highly authentic fake news. Furthermore, we propose a Shallow-Deep Multitask Learning (SDML) model for fake news, which fully uses unimodal and mutual modal features to mine the intrinsic semantics of news. Under shallow inference, we propose the momentum distillation-based light punishment contrastive learning for fine-grained uniform spatial image and text semantic alignment, and an adaptive cross-modal fusion module to enhance mutual modal features. Under deep inference, we design a two-branch framework to augment the image and text unimodal features, respectively merging with mutual modalities features, for four predictions via dedicated detection and localization projections. Experiments on both mainstream and our proposed datasets demonstrate the superiority of the model. Codes and dataset are released at https://github.com/yunan-wang33/sdml.

Figures

Figures reproduced from arXiv: 2505.06796 by the authors.

Figure 1
Figure 1. Illustrates of the news from the MFND dataset. (a) Real [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Statistics of (a) our proposed MFND dataset, and (b) comparison with 5 other multimodal fake news datasets. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Illustration of the proposed Shallow-Deep Multitask Learning (SDML) method. As for the shallow inference with green lines, we [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Ablation on layer numbers of text encoder and Contextual Aggregator. 74.56 77.83 75.45 75.67 70 72 74 76 78 80 3 4 5 6 IoUmean Score [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. D-SECURE: Dual-Source Evidence Combination for Unified Reasoning in Misinformation Detection

    cs.CV 2026-02 reject novelty 4.0 of 10

    D-SECURE fuses local manipulation detection with external evidence fact-checking, but the reported gains are undermined by a weaker strict accuracy and a post-hoc evaluation protocol.

Reference graph

Works this paper leans on

34 extracted references · 24 canonical work pages · cited by 1 Pith paper

  1. [1]

    Minigpt-v2: large language model as a unified interface for vision-language multi-task learning

    [Chen et al., 2023] Jun Chen, Deyao Zhu, Xiaoqian Shen, Xiang Li, Zechun Liu, Pengchuan Zhang, Raghuraman Krishnamoorthi, Vikas Chandra, Yunyang Xiong, and Mo- hamed Elhoseiny. Minigpt-v2: large language model as a unified interface for vision-language multi-task learning. arXiv preprint arXiv:2310.09478,

    arXiv 2023

  2. [4]

    Mini-internvl: a flexible- transfer pocket multi-modal model with 5% parameters and 90% performance

    [Gao et al., 2024] Zhangwei Gao, Zhe Chen, Erfei Cui, Yim- ing Ren, Weiyun Wang, Jinguo Zhu, Hao Tian, Shenglong Ye, Junjun He, Xizhou Zhu, et al. Mini-internvl: a flexible- transfer pocket multi-modal model with 5% parameters and 90% performance. Visual Intelligence , 2(1):1–17,

    work page 2024

  3. [6]

    [Jin et al., 2017] Zhiwei Jin, Juan Cao, Han Guo, Yongdong Zhang, and Jiebo Luo

    Curran Associates Inc. [Jin et al., 2017] Zhiwei Jin, Juan Cao, Han Guo, Yongdong Zhang, and Jiebo Luo. Multimodal fusion with recurrent neural networks for rumor detection on microblogs. In Proceedings of the 25th ACM international conference on Multimedia, pages 795–816,

    work page 2017

  4. [9]

    Bert: Pre-training of deep bidirectional transformers for lan- guage understanding

    [Kenton and Toutanova, 2019] Jacob Devlin Ming- Wei Chang Kenton and Lee Kristina Toutanova. Bert: Pre-training of deep bidirectional transformers for lan- guage understanding. In Proceedings of naacL-HLT , volume 1, page

    2019

  5. [11]

    Vilt: Vision-and-language transformer without con- volution or region supervision

    [Kim et al., 2021] Wonjae Kim, Bokyung Son, and Ildoo Kim. Vilt: Vision-and-language transformer without con- volution or region supervision. InInternational conference on machine learning, pages 5583–5594. PMLR,

    work page 2021

  6. [12]

    Edge-aware re- gional message passing controller for image forgery lo- calization

    [Li et al., 2023] Dong Li, Jiaying Zhu, Menglu Wang, Jiawei Liu, Xueyang Fu, and Zheng-Jun Zha. Edge-aware re- gional message passing controller for image forgery lo- calization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8222– 8232,

    work page 2023

  7. [14]

    Newsclippings: Automatic generation of out-of-context multimodal media

    [Luo et al., 2021] Grace Luo, Trevor Darrell, and Anna Rohrbach. Newsclippings: Automatic generation of out-of-context multimodal media. arXiv preprint arXiv:2104.05893,

    arXiv 2021

  8. [15]

    Caf-odnn: Com- plementary attention fusion with optimized deep neural network for multimodal fake news detection

    [Luvembe et al., 2024] Alex Munyole Luvembe, Weimin Li, Shaohau Li, Fangfang Liu, and Xing Wu. Caf-odnn: Com- plementary attention fusion with optimized deep neural network for multimodal fake news detection. Information Processing & Management, 61(3):103653,

    work page 2024

Show all 34 references
  1. [16]

    Detect rumors using time series of social context information on microblogging websites

    [Ma et al., 2015] Jing Ma, Wei Gao, Zhongyu Wei, Yueming Lu, and Kam-Fai Wong. Detect rumors using time series of social context information on microblogging websites. In Proceedings of the 24th ACM International on Conference on Information and Knowledge Management , CIKM ’15...

    2015

  2. [17]

    [Ma et al., 2019] Jianxin Ma, Chang Zhou, Peng Cui, Hongxia Yang, and Wenwu Zhu

    Association for Computing Machinery. [Ma et al., 2019] Jianxin Ma, Chang Zhou, Peng Cui, Hongxia Yang, and Wenwu Zhu. Learning disentangled representations for recommendation. Advances in neural information processing systems, 32,

    2019

  3. [18]

    Event-radar: Event- driven multi-view learning for multimodal fake news de- tection

    [Ma et al., 2024] Zihan Ma, Minnan Luo, Hao Guo, Zhi Zeng, Yiran Hao, and Xiang Zhao. Event-radar: Event- driven multi-view learning for multimodal fake news de- tection. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (V olume 1: Lon...

    2024

  4. [19]

    Maskfacegan: High resolution face editing with masked gan latent code optimization

    [Pernuˇs et al., 2023] Martin Pernu ˇs, Vitomir ˇStruc, and Si- mon Dobriˇsek. Maskfacegan: High resolution face editing with masked gan latent code optimization. IEEE Transac- tions on Image Processing,

    2023

  5. [20]

    Hierarchical multi- modal contextual attention network for fake news detec- tion

    [Qian et al., 2021] Shengsheng Qian, Jinguang Wang, Jun Hu, Quan Fang, and Changsheng Xu. Hierarchical multi- modal contextual attention network for fake news detec- tion. In Proceedings of the 44th international ACM SIGIR conference on research and development in information ...

    2021

  6. [21]

    Language models are unsupervised multitask learners

    [Radford et al., 2019] Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, Ilya Sutskever, et al. Language models are unsupervised multitask learners. OpenAI blog, 1(8):9,

    2019

  7. [22]

    Learning transferable visual models from nat- ural language supervision

    [Radford et al., 2021] Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agar- wal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from nat- ural language supervision. In International confere...

    2021

  8. [23]

    Detecting and grounding multi-modal media manipula- tion

    [Shao et al., 2023] Rui Shao, Tianxing Wu, and Ziwei Liu. Detecting and grounding multi-modal media manipula- tion. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 6904– 6913,

    2023

  9. [24]

    Transformer-auxiliary neural networks for image manipulation localization by operator inductions

    [Shi et al., 2023] Zenan Shi, Haipeng Chen, and Dong Zhang. Transformer-auxiliary neural networks for image manipulation localization by operator inductions. IEEE Transactions on Circuits and Systems for Video Technol- ogy, 33(9):4907–4920,

    2023

  10. [25]

    An overview of large ai models and their applications

    [Tu et al., 2024] Xiaoguang Tu, Zhi He, Yi Huang, Zhi-Hao Zhang, Ming Yang, and Jian Zhao. An overview of large ai models and their applications. Visual Intelligence, 2(1):1– 22,

    2024

  11. [26]

    Cogvlm: Visual expert for pretrained language models

    [Wang et al., 2023] Weihan Wang, Qingsong Lv, Wenmeng Yu, Wenyi Hong, Ji Qi, Yan Wang, Junhui Ji, Zhuoyi Yang, Lei Zhao, Xixuan Song, et al. Cogvlm: Visual expert for pretrained language models. arXiv preprint arXiv:2311.03079,

    2023 arXiv

  12. [27]

    Style-news: Incorporating stylized news generation and adversarial verification for neural fake news detection

    [Wang et al., 2024] Wei-Yao Wang, Yu-Chieh Chang, and Wen-Chih Peng. Style-news: Incorporating stylized news generation and adversarial verification for neural fake news detection. arXiv preprint arXiv:2401.15509,

    2024 arXiv

  13. [28]

    Modality and event adversar- ial networks for multi-modal fake news detection

    [Wei et al., 2022] Pengfei Wei, Fei Wu, Ying Sun, Hong Zhou, and Xiao-Yuan Jing. Modality and event adversar- ial networks for multi-modal fake news detection. IEEE Signal Processing Letters, 29:1382–1386,

    2022

  14. [30]

    High-resolution face swapping via latent semantics disentanglement

    [Xu et al., 2022] Yangyang Xu, Bailin Deng, Junle Wang, Yanqing Jing, Jia Pan, and Shengfeng He. High-resolution face swapping via latent semantics disentanglement. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 7642–7651,

    2022

  15. [31]

    A latent transformer for disentangled face editing in images and videos

    [Yao et al., 2021] Xu Yao, Alasdair Newson, Yann Gousseau, and Pierre Hellier. A latent transformer for disentangled face editing in images and videos. In 2021 IEEE/CVF International Conference on Computer Vision (ICCV), pages 13769–13778,

    2021

  16. [32]

    Cross-modal contrastive learning for text-to-image generation

    [Zhang et al., 2021] Han Zhang, Jing Yu Koh, Jason Baldridge, Honglak Lee, and Yinfei Yang. Cross-modal contrastive learning for text-to-image generation. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 833–842,

    2021

  17. [33]

    Multi-attentional deepfake detection

    [Zhao et al., 2021] Hanqing Zhao, Wenbo Zhou, Dongdong Chen, Tianyi Wei, Weiming Zhang, and Nenghai Yu. Multi-attentional deepfake detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2185–2194,

    2021

  18. [34]

    Exploiting context for rumour detection in social media

    [Zubiaga et al., 2017] Arkaitz Zubiaga, Maria Liakata, and Rob Procter. Exploiting context for rumour detection in social media. In Social Informatics: 9th International Conference, SocInfo 2017, Oxford, UK, September 13- 15, 2017, Proceedings, Part I 9 , pages 109–123. Springer, 2017

    2017

  19. [2015]

    Visual news: Benchmark and challenges in news image captioning

    [Liu et al., 2020] Fuxiao Liu, Yinghan Wang, Tianlu Wang, and Vicente Ordonez. Visual news: Benchmark and challenges in news image captioning. arXiv preprint arXiv:2010.03743,

    2020 arXiv

  20. [2017]

    Alias-free generative adversarial net- works

    [Karras et al., 2021] Tero Karras, Miika Aittala, Samuli Laine, Erik H ¨ark¨onen, Janne Hellsten, Jaakko Lehtinen, and Timo Aila. Alias-free generative adversarial net- works. Advances in neural information processing sys- tems, 34:852–863,

    2021

  21. [2019]

    Mvae: Multimodal variational autoencoder for fake news detection

    [Khattar et al., 2019] Dhruv Khattar, Jaipal Singh Goud, Manish Gupta, and Vasudeva Varma. Mvae: Multimodal variational autoencoder for fake news detection. In The world wide web conference, pages 2915–2921,

    2019

  22. [2020]

    Tbnet: A two- stream boundary-aware network for generic image manip- ulation localization

    [Gao et al., 2023] Zan Gao, Chao Sun, Zhiyong Cheng, Weili Guan, Anan Liu, and Meng Wang. Tbnet: A two- stream boundary-aware network for generic image manip- ulation localization. IEEE Transactions on Knowledge and Data Engineering, 35(7):7541–7556,

    2023

  23. [2021]

    Progressive growing of gans for improved quality, stability, and variation

    [Karras, 2017] Tero Karras. Progressive growing of gans for improved quality, stability, and variation. arXiv preprint arXiv:1710.10196,

    2017 arXiv

  24. [2022]

    Fusionmamba: Dynamic feature enhancement for multimodal image fusion with mamba

    [Xie et al., 2024] Xinyu Xie, Yawen Cui, Tao Tan, Xubin Zheng, and Zitong Yu. Fusionmamba: Dynamic feature enhancement for multimodal image fusion with mamba. Visual Intelligence, 2(1):37,

    2024

  25. [2023]

    The deepfake detection challenge (dfdc) dataset

    [Dolhansky et al., 2020] Brian Dolhansky, Joanna Bitton, Ben Pflaum, Jikuo Lu, Russ Howes, Menglin Wang, and Cristian Canton Ferrer. The deepfake detection challenge (dfdc) dataset. arXiv preprint arXiv:2006.07397,

    2020 arXiv

  26. [2024]

    Denoising diffusion probabilistic models

    [Ho et al., 2020] Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic models. In Proceedings of the 34th International Conference on Neural Informa- tion Processing Systems, NIPS ’20, Red Hook, NY , USA,

    2020

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.