Spatiotemporal Inconsistency Learning for DeepFake Video Detection
Open accessThe DOI matches its Crossref record; 40 of the 40 references it cites were found in a registry or an index.
Abstract
The rapid development of facial manipulation techniques has aroused public concerns in recent years. Following the success of deep learning, existing methods always formulate DeepFake video detection as a binary classification problem and develop frame-based and video-based solutions. However, little attention has been paid to capturing the spatial-temporal inconsistency in forged videos. To address this issue, we term this task as a Spatial-Temporal Inconsistency Learning (STIL) process and instantiate it into a novel STIL block, which consists of a Spatial Inconsistency Module (SIM), a Temporal Inconsistency Module (TIM), and an Information Supplement Module (ISM). Specifically, we present a novel temporal modeling paradigm in TIM by exploiting the temporal difference over adjacent frames along with both horizontal and vertical directions. And the ISM simultaneously utilizes the spatial information from SIM and temporal information from TIM to establish a more comprehensive spatial-temporal representation. Moreover, our STIL block is flexible and could be plugged into existing 2D CNNs. Extensive experiments and visualizations are presented to demonstrate the effectiveness of our method against the state-of-the-art competitors.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“However, these two-branch approaches may fragment spatiotemporal cooperation and lead to subtle artifacts being neglected.”
From TALL: Thumbnail Layout for Deepfake Video Detection (2023) · cites this paper as [15]
Citations
Cites
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318
U-Net: Convolutional Networks for Biomedical Image Segmentation
Ronneberger, O., Fischer, P., & Brox, T. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-24574-4_28 · Cited by 99,404
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946
Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset
Carreira, J., & Zisserman, A. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.502 · Cited by 9,570
Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization
Selvaraju, R. R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., & Batra, D. (2020) · International Journal of Computer Vision · doi:10.1007/s11263-019-01228-7 · Cited by 6,909
FaceForensics++: Learning to Detect Manipulated Facial Images
Rossler, A., Cozzolino, D., Verdoliva, L., Riess, C., Thies, J., & Niessner, M. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00009 · Cited by 3,314 · cited 4×
“FaceForensics++ (FF++) [26]: FF++ is a standardized benchmark for evaluation of face forgery detection methods.”
BackgroundMethodInfluentialRich Models for Steganalysis of Digital Images
Fridrich, J., & Kodovsky, J. (2012) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2012.2190402 · Cited by 2,144
Celeb-DF: A Large-Scale Challenging Dataset for DeepFake Forensics
Li, Y., Yang, X., Sun, P., Qi, H., & Lyu, S. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00327 · Cited by 1,846 · cited 1×
“Celeb-DF [15]: Celeb-DF is a large-scale dataset in DeepFake detection for developing and evaluating DeepFake detection algorithms, which comprises 5,639 high-quality DeepFake videos from YouTube and contains more than 2 million frames in total.”
MethodDeferred neural rendering: image synthesis using neural textures
Thies, J., Zollhöfer, M., & Nießner, M. (2019) · ACM Transactions on Graphics · doi:10.1145/3306346.3323035 · Cited by 1,356
Face X-Ray for More General Face Forgery Detection
Li, L., Bao, J., Zhang, T., Yang, H., Chen, D., Wen, F., & Guo, B. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00505 · Cited by 1,211
Research works printed but not linked: 3
- Lingzhi Li , Jianmin Bao , Hao Yang , Dong Chen , and Fang Wen . 2019. FaceShifter: Towards High Fidelity And Occlusion Aware Face Swapping. arXiv preprint arXiv:1912.13457 ( 2019 ). Lingzhi Li, Jianmin Bao, Hao Yang, Dong Chen, and Fang Wen. 2019. FaceShifter: Towards High Fidelity And Occlusion Aware Face Swapping. arXiv preprint arXiv:1912.13457 (2019). · arXiv
- Ji Lin , Chuang Gan , and Song Han . 2019 . TSM: Temporal Shift Module for Efficient Video Understanding. In 2019 IEEE/CVF International Conference on Computer Vision. IEEE, 7082--7092 . Ji Lin, Chuang Gan, and Song Han. 2019. TSM: Temporal Shift Module for Efficient Video Understanding. In 2019 IEEE/CVF International Conference on Computer Vision. IEEE, 7082--7092.
- Limin Wang , Zhan Tong , Bin Ji , and Gangshan Wu . 2020 a. TDN: Temporal Difference Networks for Efficient Action Recognition. CoRR , Vol. abs/ 2012 .10071 ( 2020 ). Limin Wang, Zhan Tong, Bin Ji, and Gangshan Wu. 2020 a. TDN: Temporal Difference Networks for Efficient Action Recognition. CoRR, Vol. abs/2012.10071 (2020). · arXiv
Cited by
Items 101–200 of 209
AIM-Bone: Texture Discrepancy Generation and Localization for Generalized Deepfake Detection
Liu, B., Zhang, X., Ling, H., Li, Z., Wang, R., Zhang, H., & Li, P. (2025) · IEEE Transactions on Biometrics Behavior and Identity Science · doi:10.1109/tbiom.2025.3526655 · Cited by 9
Face Swap Detection: A Systematic Literature Review
Hriez, S. (2025) · IEEE Access · doi:10.1109/access.2025.3598980 · Cited by 8
DeepFake Videos Detection via Spatiotemporal Inconsistency Learning and Interactive Fusion
Ding, X., Zhu, W., & Zhang, D. (2022) · IEEE International Conference on Sensing, Communication, and Networking (SECON) · doi:10.1109/secon55815.2022.9918605 · Cited by 7
“And their performances are in part not comparable to frame-based methods by averaging frame-level decisions to achieve the video-level score [44].”
BackgroundMethodInfluentialA Multi-Stream Fusion Approach with One-Class Learning for Audio-Visual Deepfake Detection
Lee, K., Zhang, Y., & Duan, Z. (2024) · IEEE International Workshop on Multimedia Signal Processing (MMSP) · doi:10.1109/mmsp61759.2024.10743671 · Cited by 15
“The SCNet-STIL is SCNet with STIL blocks designed to capture Spatio-Temporal Inconsis-tency [28].”
MethodExploring coordinated motion patterns of facial landmarks for deepfake video detection
Zhang, Y., Niu, R., Zhang, X., Chen, S., Wang, M., & Li, X. (2025) · Applied Soft Computing · doi:10.1016/j.asoc.2025.112974 · Cited by 9
Mining Temporal Inconsistency with 3D Face Model for Deepfake Video Detection
Cheng, Z., Chen, C., Zhou, Y., & Hu, X. (2023) · Lecture notes in computer science · doi:10.1007/978-981-99-8540-1_19 · Cited by 11
Detecting Deepfake Videos using Face Recognition and Neural Networks
Murugan, M. A., Mathu, T., & Priya, S. J. (2024) · International Conference on Cognitive Robotics and Intelligent Systems (ICC-ROBINS) · doi:10.1109/icc-robins60238.2024.10534025 · Cited by 7
“To capture temporal dependencies[12] within the sequence of frames, the model incorporates an LSTM layer.”
MethodDistilling Multi-Level Semantic Cues Across Multi-Modalities for Face Forgery Detection
Yu, L., Xie, T., Liu, C., Jin, G., Ding, Z., & Xie, H. (2024) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2024.3524602 · Cited by 5
“Consequently, there is a growing number of studies focusing on temporal incoherence [35], [36], [37], [38].”
GFADE: generalized feature adaptation and discrimination enhancement for deepfake detection
Tian, Z., & Yi, J. (2025) · PeerJ Computer Science · doi:10.7717/peerj-cs.2879 · Cited by 6
“For instance, Face X-ray ( Li et al., 2020a ) uses the boundaries between altered faces and their backgrounds to identify spatial inconsistencies, while Local relation learning (LRL) ( Chen et al., 2021 ) combines RGB image analysis with frequency-aware cues to enhance detection capabilities.”
Peng, C., Chen, Y., Liu, D., Wang, N., & Gao, X. (2025) · Neural Networks · doi:10.1016/j.neunet.2025.107899 · Cited by 6
Self-citationKukanov, I., & Ng, J. W. (2025) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3746027.3761982 · Cited by 9
Hybrid Domain Meta-Learning Network for Face Forgery Detection and Localization in Deepfakes
Zhao, H., Liu, B., Hu, Y., Li, J., & Li, C. (2023) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn54540.2023.10191323 · Cited by 8
Self-citationExploiting Inconsistencies in Object Representations for Deepfake Video Detection
Bhaumik, K. K., & Woo, S. S. (2023) · doi:10.1145/3595353.3595885 · Cited by 5
3D data augmentation and dual-branch model for robust face forgery detection
Zhou, C., Li, F. W. B., Song, C., Dong, Z., & Yang, B. (2025) · Graphical Models · doi:10.1016/j.gmod.2025.101255 · Cited by 7
Bi-Stream Coteaching Network for Weakly-Supervised Deepfake Localization in Videos
Li, Z., Teng, Z., Zhang, B., & Fan, J. (2025) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2025.3533906 · Cited by 9
“Moreover, we compare our IPM with other inconsistency-based methods by replacing it with STIL [34] and DFDC [90], as presented in the first three rows of Table VIII.”
InfluentialDeepfake Video Detection Based on the Decomposition of Spatial-Temporal Attention Mechanism in ViViT
Sun, G., & Lian, Z. (2024) · IEEE International Symposium on Parallel and Distributed Processing with Applications (ISPA) · doi:10.1109/ispa63168.2024.00221 · Cited by 6
“In Gu's study [17], the STIL block was proposed as a method to capture temporal inconsistencies across two perpendicular directions.”
Visual Explanations for Exposing Potential Inconsistency of Deepfakes
Pei, P., Zhao, X., Cao, Y., & Hu, C. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-25115-3_5 · Cited by 5
Dual-Modality Co-Learning for Unveiling Deepfake in Spatio-Temporal Space
Guan, J., Zhou, H., Guo, Z., Hu, T., Deng, L., Quan, C., Fang, M., & Zhao, Y. (2023) · International Conference on Multimedia Retrieval · doi:10.1145/3591106.3592284 · Cited by 5
“When CNN-based methods [21, 61] downsample features, they tend to focus on the dominant motion, such as mouth movements and blinking, while minor distortions, such as artifacts on the edges of cheeks resulting from imperfect face-swapping blending, may be overlooked.”
BackgroundInfluentialPark, C. S., Muneer, M. S., & Woo, S. S. (2025) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3746252.3760853 · Cited by 5
“We compare against representative baselines from two categories: visual uni-modal methods [12, 13, 24, 34, 40], and multimodal approaches [9, 28, 39, 41].”
DeepShield: Fortifying Deepfake Video Detection with Local and Global Forgery Analysis
Cai, Y., Li, J., Li, Z., Chen, W., Lan, R., Xie, X., Luo, X., & Li, G. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.01164 · Cited by 18
Self-citationKlein, N., Tak, H., Fullwood, J., Regmi, K., Spinoulas, L., Sivaraman, G., Chen, T., & Khoury, E. (2025) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3746027.3761981 · Cited by 8
Robust Deepfake Detection via Perturbation Domain Alignment
Lu, L., Wang, Y., Zhang, L., & Guo, Y. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10890870 · Cited by 5
“CD2 DFDC DFDCP Face X-ray [31] 79.50 65.50 80.92 STIL [32] 75.58 - - LRL [12] 78.26 - 76.53 FTCN [33] 86.90 71.00 74.0 Two-branch [34] 76.65 - - DAM [35] 75.30 - 72.80 LipForensics [14] 82.”
InfluentialWang, Y., Sun, Q., Zhang, J., Rong, D., Shen, C., & Wang, X. (2025) · Information Fusion · doi:10.1016/j.inffus.2025.103708 · Cited by 9
Detection of Compressed DeepFake Video Drawbacks and Technical Developments
Humidan, A.-S., Abdullah, L. N., & Halin, A. B. A. (2022) · International Conference on Signal Processing and Information Security (ICSPIS) · doi:10.1109/icspis57063.2022.10002433 · Cited by 7
“Another spatio-temporal model was proposed in [40] that extracts the temporal inconsistency between consecutive frames in both horizontal and vertical directions.”
BackgroundMining collaborative spatio-temporal clues for face forgery detection
Ding, B. Y., Fan, Z., Zhao, Z., & Xia, S. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-16173-4 · Cited by 7
“These features are inconsistent between adjacent frames of the video, including eye blinks [31], facial movements [49], lip-reading [24], optical flow [2, 6, 46], and texture inconsistency [20, 21, 59].”
Interpretable DeepFake Detection Based on Frequency Spatial Transformer
Luan, T., Liang, G., & Pei, P. (2024) · International Journal of Emerging Technologies and Advanced Applications · doi:10.62677/ijetaa.2402108 · Cited by 5
“The state-of-the-art methods for comparison are Xception[22], DILNet[28], Grad-CAM[29], DIANet[30], STIL[31], FInter[32], and ViTHash [33].”
MethodLai, C.-Y., Jian, C.-Y., Chuang, P.-C., Lee, C.-M., Hsu, C., Chiou-Ting, H., & Lin, C. (2026) · International Journal of Computer Vision · doi:10.1007/s11263-025-02606-0 · Cited by 3
Hyun, J., & Kim, H. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-37429-5_29 · Cited by 3
Mi, Z., Jiang, X., Sun, T., Xu, K., Xu, Q., & Meng, L. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-5606-3_23 · Cited by 4
Frame-Level Temporal Difference Learning for Partial Deepfake Speech Detection
Li, M., Zhang, X.-P., & Zhao, L. Y. (2025) · IEEE Signal Processing Letters · doi:10.1109/lsp.2025.3592111 · Cited by 9
“To address these challenges, we propose a Temporal Difference Attention Module (TDAM) that captures unnatural temporal fluctuations across utterances rather than relying on explicit boundary detection.”
Local similarity anomaly for general face forgery detection
Dai, Y., Fei, J., Xia, Z., Liu, J.-N., & Weng, J. (2023) · Journal of Image and Graphics · doi:10.11834/jig.221006 · Cited by 3
Partial Reconstruction Error for Deepfake Detection
Zhang, Y., Meng, Z., Peng, B., Dong, J., Chu, B., & Wang, W. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10889075 · Cited by 4
“FF++ Celeb-DF DiFF STIL [10] (MM’21) None 91.25 70.45 51.43 CoReD [15] (MM’21) 94.65 70.48 60.48 LTW [24] (AAAI’21) 89.64 64.10 58.48 S-prompts [25] (NIPS’22) 82.48 70.81 58.71 SLADD [3] (CVPR’22) 95.48 70.84 64.24 IDC [18] (ICIP’23) 97.54 64.25 58.64 DFIL [21] (MM’23) 84.77 67.62 59.18 CADDM [8]…”
InfluentialDeepfake detection in videos with multiple faces using geometric-fakeness featuresPreprint
Vyshegorodtsev, K., Kudiyarov, D., Balashov, A. K., & Kuzmin, A. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2410.07888 · Cited by 6
C3D-VIT: Consistency-Aware 3D Vision Transformer for Face Forgery Detection
Zhang, J., Zhang, P., & Wang, J. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10887985 · Cited by 2
Frequency-Assisted Temporal Upsampling Artifacts Representation Learning for Face Forgery Detection
Sun, R., Yu, X., Wang, F., Da, Z., Zhang, Y., & Gao, J. (2025) · IEEE Transactions on Biometrics Behavior and Identity Science · doi:10.1109/tbiom.2025.3569645 · Cited by 2
“STIL [26] block utilizes temporal difference within adjacent frames along two orthogonal directions to capture fine-grained temporal incoherence.”
InfluentialChasing Shadows: Solving Deepfake Detection Benchmarks Using Irrelevant Features Only
Xu, Y., Terhörst, P., Pedersen, M., & Raja, K. B. (2025) · International Conference on Automatic Face and Gesture Recognition · doi:10.1109/fg61629.2025.11099163 · Cited by 2
“Compared to the extensive research on Deepfake detection [37], [40], [4], [18], [18], [40], [46], [14], [15], [47], [24], the analysis of biases within the realm of Deepfake detection remains relatively underexplored.”
Wang, Y., Sun, Q., & Rong, D. (2025) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3746027.3755455 · Cited by 5
“Earlydeepfakedetection methods[14,22,33,35,47]haveachieved significant progress in unimodal videos, primarily focusing on visual artifacts caused by local texture inconsistencies and temporal artifacts across frames.”
FakeRadar: Probing Forgery Outliers to Detect Unknown Deepfake Videos
Li, Z., Li, J., Cai, Y., Chen, J., Luo, X., Li, G., & Lan, R. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.01243 · Cited by 7
“Other video detection studies [18, 39, 62] incorporate both temporal and spatiotemporal features to establish a more comprehensive detection framework.”
Self-citationDomain-Adaptive Representation Learning for Multimodal Deepfake Detection
Haq, I. U., & Malik, K. M. (2025) · IEEE Transactions on Computational Social Systems · doi:10.1109/tcss.2025.3637242 · Cited by 3
“To demonstrate the effectiveness of our approach, we compare its re-sults against state-of-the-art deepfake detection techniques, including single-modality (visual) methods [36], [64], [65], [66], [67], [68] and multimodal approaches [16], [17], [19], [54], [63], [69], [70], [71].”
Dual-Granularity Contrastive Learning for DeepFake Detection
Zhang, F. X., Shao, C., Du, K., Guo, Y., Song, P., Cao, L., & Yuan, X. (2026) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2026.3653569 · Cited by 2
“Moreover, Gu et al. proposed spatial-temporal inconsistency learning (STIL) [33] for forgery video detection, where the temporal inconsistency module, spatial inconsistency module, and spatial-temporal information complementary module were designed to enable the model to have more comprehensive…”
AVIR: Audio-Visual Information Refinement for Robust Multimodal Deepfake Detection
Liu, X., Yang, Y. Y., Li, X., Zhao, G., & Zhao, Y. (2026) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2026.3691216 · Cited by 2
“Thus, many researches have been proposed to detect Deepfake, such as frequency clues [2], style mixture [3], multi-task learning [4] and facial embedding discrepancy [5].”
Trident : Detecting Face Forgeries With Domain-Adversarial Triplet Learning
Kara, M. H., Dündar, A., & Güdükbay, U. (2026) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2026.3703428 · Cited by 3
“To capture these cues, Gu et al. [18] proposed spatiotemporal inconsistency learning, while Amerini et al. [19] utilized optical-flow convolutional neural networks (CNNs) to uncover compression-robust artifacts.”
Towards Spatio-temporal Collaborative Learning: An End-to-End Deepfake Video Detection Framework
Guo, W., Du, S., Deng, H., Yu, Z., & Feng, L. (2023) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn54540.2023.10191479 · Cited by 2
“Therefore, many researchers recently presented advanced video-based methods [11]–[15], [18] to capture temporal forgery traces as discernible features to detect forged video.”
BackgroundMethodInfluentialMVFNet: Multipurpose Video Forensics Network using Multiple Forms of Forensic Evidence
Nguyen, T. D., & Stamm, M. C. (2025) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv61041.2025.00221 · Cited by 5
“These multi-scale, spatially distributed decisions are then aggregated and used to produce an overall pre-training loss: where c ∗ is the true source camera model, 1 ( · ) is the indicator function and λ is the relative weight of each scale.”
Chameleon: Benchmarking Detection and Backtracking on Commercial-Grade AI-Generated Videos
Liao, X., Zeng, M., Chen, C. E., Lin, N., Wang, Z., & Yang, A. (2026) · International Conference on Multimedia Retrieval · doi:10.1145/3805622.3810862 · Cited by 3
Attention-guided Fine-grained Feature Learning For Robust Face Forgery Detection
Zhu, Y., Gao, J., Liu, Q., & Zhou, X. (2022) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr56361.2022.9956159 · Cited by 2
“Many significant works ([6], [13], [7], [5], [20], [30], [10], [9]) so far have been dedicated considerable efforts to face forgery detection.”
BackgroundDeepFake Doctor: Diagnosing and Treating Audio-Video Fake DetectionPreprint
Klemt, M., Segna, C., & Rohrbach, A. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2506.05851 · Cited by 3
“(e) Popular datasets, such as FakeAVCeleb, seem to be fairly saturated [36, 18, 19, 17].”
Local Corner and Motion Key Point Trajectory Extraction for Facial Forgery Identification
Liu, Q., & Xue, Z. (2023) · Security and Communication Networks · doi:10.1155/2023/1648912 · Cited by 2
“Te test results are shown in Table 5, Xception [18], LRNet [11], DSP-FWA [24], Capsule [25], Single XceptionNet [26], FWA [7], LipForensics [31], STIL [33], ADDNet-3D [34], and ours are compared.”
Survey on generative visual media detection and security
Duan, J., Liu, S., Guan, J., Jie, C. A. O., & He, R. (2025) · Scientia Sinica Informationis · doi:10.1360/ssi-2024-0354 · Cited by 1
HTMDF-DD: Hybrid triple modality based spatial–temporal features early fusion for deepfake detection
Das, A. K., & Naskar, R. (2025) · Computers & Electrical Engineering · doi:10.1016/j.compeleceng.2025.110860 · Cited by 3
Bhatnagar, A., Pandey, R. C., Choudhary, V., & Singh, A. (2025) · IEEE International Conference on Advances in Computing Research on Science Engineering and Technology (ACROSET) · doi:10.1109/acroset66531.2025.11280626 · Cited by 2
Face Forgery Detection With CLIP-Enhanced Multi-Encoder Distillation
Peng, C., Yan, T., Liu, D., Wang, N., Hu, R., & Gao, X. (2025) · IEEE Transactions on Image Processing · doi:10.1109/tip.2025.3644125 · Cited by 1
“STIL [8] model proposes the TIM temporal modeling paradigm, which deeply mines the temporal di ff erence information of adjacent frames in the video in the horizontal and vertical dimensions, and enhances the model’s ability to perceive the temporal coherence of forged videos.”
Zhang, R., Li, B., & Xu, D. (2026) · Symmetry · doi:10.3390/sym18020259 · Cited by 1
STGT: Spatio-temporal graph transformer for robust deepfake detection using dynamic facial features
Pandey, A., & Solanki, A. (2026) · Signal Image and Video Processing · doi:10.1007/s11760-026-05191-y · Cited by 1
Asuai, C., Ogbogbo, G. O., Hosni, H., & Khan, M. I. (2026) · International Journal of Wireless and Microwave Technologies · doi:10.5815/ijwmt.2026.02.10 · Cited by 1
Efficient deepfake detection: a hybrid lightweight framework for cross-domain generalization
Nagar, H. V., Gandhi, V., & Rathod, D. M. (2026) · Signal Image and Video Processing · doi:10.1007/s11760-026-05419-x · Cited by 1
Shi, S., Tan, H., Tan, Z., Feng, S., Liu, A., Escalera, S., & Wan, J. (2026) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3767308.3836508 · Cited by 1
Embodied Intelligence Security with Vision-language Models: A Survey
Duan, J., Zhu, H., Liu, C., Li, S. Y., Tang, X., & Wang, Y. (2026) · Machine Intelligence Research · doi:10.1007/s11633-025-1626-x · Cited by 2
Hybrid Spatio-Temporal Network for Face Forgery Detection
Liu, X., Gao, S., Zhou, P., Liu, J., Luo, X., Zhang, L., & Zhang, B. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-47665-5_21 · Cited by 2
Dual-PST: Dual-Branch SpatioTemporal-Planar Network for Video Forgery Detection
Liu, S., Zhang, Z., Duan, J., Cao, J., & Zheng, A. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10890154 · Cited by 1
“Although some studies have employed spatiotemporal neural networks to effectively detect temporal inconsistency in videos [7], [8], these methods primarily focus on spatial feature extraction within individual frames and fail to effectively incorporate temporal relationships into local feature…”
Robust deepfake detection using Long Short-Term Memory networks for video authentication
Surapaneni, R. K., Syed, H., Kakarala, H., & Yaragudipati, V. S. S. (2025) · Informatyka Automatyka Pomiary w Gospodarce i Ochronie Środowiska · doi:10.35784/iapgos.6777 · Cited by 3
Research on robustness evaluation and verification for facial deepfake detection
Liu, S., Chen, J.-R., Chen, W., & Li, J. (2023) · International Conference on Compute and Data Analysis (ICCDA) · doi:10.1145/3629264.3629268
“Other methods often combine time domain features for detection [12] [14] ,which can find the incoherence of video through time information.”
MethodSelf-citationUstvarjanje ponarejenih videoposnetkov s pomočjo difuzijskih modelov
Markelj, B., Peer, P., & Batagelj, B. (2024) · Računalniška obdelava slik in njena uporaba v Sloveniji (ROSUS) · doi:10.18690/um.feri.1.2024.8
A Temporal Consistency Learning Framework for Face Forgery Detection
Wang, X., Zhu, F., Li, L., & Tan, X. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-4399-5_21
DFADNet: A Diverse-Feature Adaptive Network for Web3.0-Oriented Deep Forgery Detection
Chen, G., Liu, Y., Ye, Y., & Lei, T. (2024) · Communications in computer and information science · doi:10.1007/978-981-97-9412-6_17
DeepFake Detection Using ResNeXt, LSTM And ESRGANPreprint
A., D. S., Bibhuti, N. K., Agarwal, D., & Gupta, I. J. V. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.5028609 · Cited by 1
Deepfake Video Detection Guided by Identity and Temporal Inconsistency
Zhang, Y., Peng, B., Dong, J., You, W., & Wang, W. (2025) · Lecture notes in computer science · doi:10.1007/978-981-96-1068-6_18
Survey on Technologies of Video Deepfake Detection
Li, C., Song, Z., Liang, Y., & Wang, X. A. (2025) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-3-031-86149-9_18
Dual-Domain Reconstruction Information Guided Two-Stage Deepfake DetectionPreprint
Zhang, F., Peng, J., Shao, C., Cao, L., Du, K., Guo, Y., Tian, S., & Song, P. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5279098
Deepfake detection based on Spatio-Temporal Fusion
Chen, T., Hu, J., & Yang, S. (2025) · International Conference on Information Management and Computing Technology (ICIMCT) · doi:10.1109/icimct.2025.00024
“The experiments select mainstream forged face detection models, including Xception [44], Face X-ray [45], and STIL [12], as comparison baselines.”
MyGO: Modality-incomplete Fake News Video Detection via Prompt-assisted Modality Disentangling Model
Qiu, M., Tan, Z., & Bao, B. (2026) · ACM Transactions on Multimedia Computing Communications and Applications · doi:10.1145/3785481
“Moreover, SVFEND [29] designs cross-modality transformer encoders to incorporate information from different modalities.”
Markelj, B. (2023) · Repository of the University of Ljubljana (University of Ljubljana) · Cited by 2
Constructing Audio-Visual Heterogeneous Graphs for Deepfake Detection
Xie, N., Li, J., Deng, Y., Guo, M., & Yang, L. (2025) · Signals and communication technology · doi:10.1007/978-3-031-94973-9_4
Self-citationShah, K., Shah, S., & Ajwalia, M. (2025) · International Conference on IoT Based Control Networks and Intelligent Systems (ICICNIS) · doi:10.1109/icicnis66685.2025.11315573
“Combining spatial CNNs with temporal models like LSTMs can improve video-level detection by capturing both frame artifacts and temporal inconsistencies [13].”
Advancing Deepfake Detection Through Head Pose Estimation on New-Generation Datasets
Bisogni, C., Castiglione, A., Migliaccio, M., Miglino, A., Nappi, M., & Pero, C. (2026) · Lecture notes in computer science · doi:10.1007/978-981-95-3551-4_24
Kumar, P. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6052494
“Advanced deep learning techniques are used to create deepfakes, and people looking to trick and control others increasingly turn to recurrent neural networks, autoencoders, and Generative Adversarial Networks (GANs) (Gu et al., 2021).”
SADN: Saliency-Based Attention Discrimination Network for Deepfake Detection
Wang, X., Lu, D., Dong, J., Hedjam, R., & Zhong, G. (2026) · Lecture notes in computer science · doi:10.1007/978-981-95-5696-0_8
DeepFake Analysis : A Review of Recent Studies, Detection Methods and Trends in Deepfake Detection
Kaur, P., & Sharma, C. (2025) · International Conference on Innovations and Emerging Technologies in AI & Communication Systems (IETACS) · doi:10.1109/ietacs68750.2025.11385353
3D differential decomposition for video deepfake detection with identity suppression
Gao, J., Micheletto, M., Orrù, G., Feng, X., & Marcialis, G. L. (2026) · Signal Processing Image Communication · doi:10.1016/j.image.2026.117525
Adaptive Consensus Multi-Teacher Distillation for Generalizable Face Forgery Detection
Jing, J., Sun, F., Peng, C., Liu, D., Wang, N., & Gao, X. (2026) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2026.3668432
Integrating Vision Transformer and MesoNet for Enhanced Deepfake Detection in Videos
Mandal, G., & Kumar, G. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-19242-4_3
Are DeepFakes Realistic Enough? Exploring Semantic Mismatch as a Novel Challenge
Deshmukh, S. N., Hambarde, K., Costa, J. C., Proença, H. P. M. C., & Roxo, T. (2026) · IEEE Access · doi:10.1109/access.2026.3730281
“Early methods relied on uni-modal analysis through CNNs and spatiotemporal networks trained on established benchmarks [25, 14], methods exploiting lip dynamics, irregular blinking, and physiological inconsistencies [20, 16, 15], and spectral approaches for synthetic speech detection [2].”
QIN, L., Feng, C., Zhang, Z., & Chen, H. (2026) · Image and Vision Computing · doi:10.1016/j.imavis.2026.106016
Open-Unfairness Adversarial Mitigation for Generalized Deepfake Detection
Li, Z. G., Teng, Z., Zhang, B., & Fan, J. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.00073 · Cited by 3
“These methods exploit various cues, such as frequency-based analysis [21, 23, 32] and inconsistency detection [11, 43], among others.”
Explainable Deepfake Video Detection Using Visual, Temporal, Physiological, and Audio-Visual Cues
Bhatia, A., Ibrahim Mohamed, S., Gokulsivaraj, V. P., & Krishnan, K. (2026) · International Conference on Recent Advances in Electrical, Electronics, Ubiquitous Communication, and Computational Intelligence (RAEEUCCI) · doi:10.1109/raeeucci67649.2026.11504825
“Gu et al. [23] explored spatiotemporal features using 3D CNNs for video-level deepfake detection, showing that temporal artifacts provide complementary information to spatial cues.”
Li, J., Liao, G., Wang, Y., Liu, X., & Liu, B. Q. (2026) · Mathematics · doi:10.3390/math14101612
Self-citationDeMorph: Explainable Multi-Forensic Single-Image Detection of Deepfake and Morphed Faces
Abhinayapriya, P. C., Hema Varshini, R., & Suganya, N. (2026) · International Conference on Augmented Intelligence and Sustainable Systems (ICAISS) · doi:10.1109/icaiss68683.2026.11526607
When Deepfake Detection Meets Graph Neural Network: A Unified and Lightweight Framework
Liu, H., Gong, C., He, M., Li, J., Han, K., & Luo, S. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770854.3780219 · Cited by 1
Self-citationICR-Net: Robust Deepfake Detection Under Temporal Corruption
Park, C., Choi, H., Muneer, M. S., Le, B. M., & Woo, S. S. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-1465-5_24 · Cited by 2
Qiu, M.-X., Chuang, K., & Lai, S. (2026) · International Conference on Automatic Face and Gesture Recognition · doi:10.1109/fg67764.2026.11556934
“To address the limitations of spatial and temporal detectors, spatial-temporal models [53], [54], [55], [17], [18] have been developed to analyze both cues jointly.”
EchoFence: Non-Intrusive Forgery Detection in Video Conferencing via Ultrasonic Sensing
Zhao, L., Shi, L., Liu, J., Xiao, R., & Han, J. (2026) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom59046.2026.11571188
“It performs local verification on the ultrasonic responses and video streams to detect such video forgery attacks and ensure authenticity. anomalies [7], [8], and lip-audio mismatch [9], [10].”
Check-Agent: Multimodal fraud detection via a multi-agent framework
Zhao, Y., Xu, Z., & Wu, X. (2026) · Knowledge-Based Systems · doi:10.1016/j.knosys.2026.116511
Khan, M. A., Madhusudhan, H. S., & Gupta, P. (2026) · Procedia Computer Science · doi:10.1016/j.procs.2026.06.045
Maulana, M. F., & Poetro, B. S. W. (2026) · Jurnal Rekayasa Sistem Informasi dan Teknologi · doi:10.70248/jrsit.v3i4.3840
Sun, R., Zhang, Y., Yu, X., Li, M., & Gao, J. (2026) · IEEE Transactions on Dependable and Secure Computing · doi:10.1109/tdsc.2026.3712721 · Cited by 1
“STIL [28] block utilizes temporal difference within adjacent frames along two orthogonal directions to capture fine-grained temporal incoher-ence.”
A novel temporal attention regularization loss for robust and explainable deepfake detection
Alrawahneh, A. A.-M., Abdullah, S. R. S., Wahab, A. N. A., Taylor, S. K., Rahim, N. R. N. A., & Kok, V. J. (2026) · Discover Artificial Intelligence · doi:10.1007/s44163-026-01784-1
HFVideoSwin: High-Frequency Spatio-Temporal Features for More Generalizable Deepfake Video Detection
Atamna, M., Tkachenko, I., & Miguet, S. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-31452-9_11
Generalizable Deepfake Detection via Simplicity-Bias-Aware CLIP Adaptation
Yahchouchi, C., Roggero, N., Saroul, L., & Dantcheva, A. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-31438-3_38 · Cited by 1
Compressed-Domain Spatio-Temporal Forensics for AI-Generated Video Detection
Mu, W.-P., Han, X., Cui, H., Xu, Q., Ba, Z.-J., & Sun, T.-F. (2026) · 2026 IEEE 3rd International Conference on Computer Vision and Deep Learning (DLCV) · doi:10.1109/dlcv69906.2026.11635513
“Existing methods mainly operate in the pixel domain, frequency-domain clues, temporal inconsistencies, or multi-modal mismatch for forensic analysis [13]–[15].”
MAS Optimization for Temporal Deepfake Detection: Reducing Cross-Fold Variance in Video Forensics
Nguyen, M. N., & Thuân, N. Đ. (2026) · SN Computer Science · doi:10.1007/s42979-026-05318-5
“More recent approaches incorporate attention mecha - nisms to enhance temporal feature aggregation and empha - size informative frames or regions across time [ 11, 12].”
Metadata
- Authors
- Gu, Z., Chen, Y., Yao, T., Ding, S., Li, J., Huang, F., & Ma, L.
- Affiliations
- Shanghai Jiao Tong University (CN)
- Tencent (China) (CN)
- Venue
- Proceedings of the 29th ACM International Conference on Multimedia
- Publisher
- ACM
- Published
- 2021-10-17
- Pages
- 3473-3481
- OpenAlex
- W3207531490
- Cited by
- 189
- References
- 36
- Citations by year
- 2026: 36 · 2025: 53 · 2024: 37 · 2023: 44 · 2022: 19
- FWCI
- 8.18
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- ACM International Conference on Multimedia (ACM MM) (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- National Key Research and Development Program of China (No. 2019YFC1521104)
- National Natural Science Foundation of China (61972157)
- Shanghai Municipal Science and Technology Major Project (2021SHZDZX0102)
- Zhejiang Lab (2020NB0AB01)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Digital Media Forensic Detection
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Generative Adversarial Networks and Image Synthesis
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Video Quality Assessment
- Keywords
- deepfake detection, video forgery detection, face manipulation, spatial inconsistency, time inconsistency, temporal modeling, 2D convolutional neural network
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.