CMTA: Leveraging Cross-Modal Temporal Artifacts for Generalizable AI-Generated Video Detection

• URL: http://arxiv.org/abs/2605.00630v1
• Date: Fri, 01 May 2026 13:04:14 GMT
• Title: CMTA: Leveraging Cross-Modal Temporal Artifacts for Generalizable AI-Generated Video Detection
• Abstract summary: Video synthesis techniques pose an unprecedented challenge to digital authenticity.<n>We identify a distinctive fingerprint in cross-modal temporal artifact (CMTA)<n>We propose the CMTA framework, a cross-modal detection approach that captures these unique temporal artifacts.
• Score: 45.739302264021795
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The proliferation of advanced AI video synthesis techniques poses an unprecedented challenge to digital video authenticity. Existing AI-generated video (AIGV) detection methods primarily focus on uni-modal or spatiotemporal artifacts, but they overlook the rich cues within the visual-textual cross-modal space, especially the temporal stability of semantic alignment. In this work, we identify a distinctive fingerprint in AIGVs, termed cross-modal temporal artifact (CMTA). Unlike real videos that exhibit natural temporal fluctuations in cross-modal alignment due to semantic variations, AIGVs display unnaturally stable semantic trajectories governed by given input prompts. To bridge this gap, we propose the CMTA framework, a cross-modal detection approach that captures these unique temporal artifacts through joint cross-modal embedding and multi-grained temporal modeling. Specifically, CMTA leverages BLIP to generate frame-level image captions and utilizes CLIP to extract corresponding visual-textual representations. A coarse-grained temporal modeling branch is then designed to characterize temporal fluctuations in cross-modal alignment with a GRU. In parallel, a fine-grained branch is constructed to capture intricate inter-frame variations from integrated visual-textual features with a Transformer encoder. Extensive experiments on 40 subsets across four large-scale datasets, including GenVideo, EvalCrafter, VideoPhy, and VidProM, validate that our approach sets a new state-of-the-art while exhibiting superior cross-generator generalization. Code and models of CMTA will be released at https://github.com/hwang-cs-ime/CMTA

Related papers

• Video as Natural Augmentation: Towards Unified AI-Generated Image and Video Detection [17.82437958396922]
  We show that SOTA AI-generated image detectors often collapse when applied to frames extracted from videos.<n>Motivated by these findings, we propose VINA, a unified AIGC detection framework that jointly trains on image and video data.
  arXiv  Detail & Related papers  (2026-05-21T04:11:32Z)
• Towards multi-modal forgery representation learning for AI-generated video detection and localization [8.396379868583718]
  AI-generated videos pose escalating risks of semantic distortion and misuse.<n>Most existing AI-generated video detectors remain limited by single- or partial-modality of data modeling.<n>We introduce a core architecture that integrates an LMM semantic branch with atemporal video (ST) visual branch and a multi-scale partial-spoof (PS) audio branch.<n>This multi-modal approach enables simultaneous detection and fine-grained temporal localization of partially manipulated AI-generated video forgeries.
  arXiv  Detail & Related papers  (2026-05-08T04:35:50Z)
• ATSS: Detecting AI-Generated Videos via Anomalous Temporal Self-Similarity [44.13741922118129]
  We identify a distinctive fingerprint in AIGVs, termed anomalous temporal self-similarity (ATSS)<n>We propose ATSS, a multimodal detection framework that exploits this insight via a triple-similarity representation and a cross-attentive fusion mechanism.<n>ATSS significantly outperforms state-of-the-art methods in terms of AP, AUC, and ACC metrics.
  arXiv  Detail & Related papers  (2026-04-05T09:10:13Z)
• CtrlVDiff: Controllable Video Generation via Unified Multimodal Video Diffusion [62.04833878126661]
  We tackle the dual challenges of video understanding and controllable video generation within a unified diffusion framework.<n>We propose CtrlVDiff, a unified diffusion model trained with a Hybrid Modality Control Strategy (HMCS) that routes and fuses features from depth, normals, segmentation, edges, and graphics-based intrinsics (albedo, roughness, metallic)<n>Across understanding and generation benchmarks, CtrlVDiff delivers superior controllability and fidelity, enabling layer-wise edits (relighting, material adjustment, object insertion) and surpassing state-of-the-art baselines while remaining robust when some modalities are unavailable.
  arXiv  Detail & Related papers  (2025-11-26T07:27:11Z)
• Complementary and Contrastive Learning for Audio-Visual Segmentation [74.11434759171199]
  We present Complementary and Contrastive Transformer (CCFormer), a novel framework adept at processing both local and global information.<n>Our method sets new state-of-the-art benchmarks across the S4, MS3 and AVSS datasets.
  arXiv  Detail & Related papers  (2025-10-11T06:36:59Z)
• BindWeave: Subject-Consistent Video Generation via Cross-Modal Integration [56.98981194478512]
  We propose a unified framework that handles a broad range of subject-to-video scenarios.<n>We introduce an MLLM-DiT framework in which a pretrained multimodal large language model performs deep cross-modal reasoning to ground entities.<n>Experiments on the OpenS2V benchmark demonstrate that our method achieves superior performance across subject consistency, naturalness, and text relevance in generated videos.
  arXiv  Detail & Related papers  (2025-10-01T02:41:11Z)
• LoViC: Efficient Long Video Generation with Context Compression [68.22069741704158]
  We introduce LoViC, a DiT-based framework trained on million-scale open-domain videos.<n>At the core of our approach is FlexFormer, an expressive autoencoder that jointly compresses video and text into unified latent representations.
  arXiv  Detail & Related papers  (2025-07-17T09:46:43Z)
• FancyVideo: Towards Dynamic and Consistent Video Generation via Cross-frame Textual Guidance [3.6519202494141125]
  We introduce FancyVideo, an innovative video generator that improves the existing text-control mechanism.<n>CTGM incorporates the Temporal Information (TII) and Temporal Affinity Refiner (TAR) at the beginning and end of cross-attention.<n>Our approach achieves state-of-the-art T2V generation results on the EvalCrafter benchmark.
  arXiv  Detail & Related papers  (2024-08-15T14:47:44Z)
• Free-Bloom: Zero-Shot Text-to-Video Generator with LLM Director and LDM Animator [59.589919015669274]
  This study focuses on zero-shot text-to-video generation considering the data- and cost-efficient. We propose a novel Free-Bloom pipeline that harnesses large language models (LLMs) as the director to generate a semantic-coherence prompt sequence. We also propose a series of annotative modifications to adapting LDMs in the reverse process, including joint noise sampling, step-aware attention shift, and dual-path.
  arXiv  Detail & Related papers  (2023-09-25T19:42:16Z)
• UMMAFormer: A Universal Multimodal-adaptive Transformer Framework for Temporal Forgery Localization [16.963092523737593]
  We propose a novel framework for temporal forgery localization (TFL) that predicts forgery segments with multimodal adaptation. Our approach achieves state-of-the-art performance on benchmark datasets, including Lav-DF, TVIL, and Psynd.
  arXiv  Detail & Related papers  (2023-08-28T08:20:30Z)
• Spatio-Temporal Transformer for Dynamic Facial Expression Recognition in the Wild [19.5702895176141]
  We propose a method for capturing discnative features within each frame model. We utilize the CNN to translate each frame into a visual feature sequence. Experiments indicate that our method provides an effective way to make use of the spatial and temporal dependencies.
  arXiv  Detail & Related papers  (2022-05-10T08:47:15Z)

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.