Research map / Vision-language and multimodal / Cross-modal learning
X-AVDT: Audio-Visual Cross-Attention for Robust Deepfake Detection
Youngseo Kim, Kwan Yun, Seokhyeon Hong, Sihun Cha, Colette Suhjung Koo, Junyong Noh
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add audio-visual cross-attention feature psi (ablation) | X-AVDT · MMDF average | AUROC: 88.22 -> 95.29 | Improved |
| add video composite phi (inversion-induced discrepancies) (ablation) | X-AVDT · MMDF average | AUROC: 90.21 -> 95.29 | Improved |
| add reconstruction residual channel (ablation) | X-AVDT · MMDF average | AUROC: 93.82 -> 95.29 | Improved |
| add triplet loss to BCE (ablation) | X-AVDT · MMDF average | AUROC: 92.64 -> 95.29 | Improved |
| use cross-attention instead of temporal attention features (ablation) | X-AVDT · MMDF | AUROC: 83.92 -> 91.56 | Improved |
| use cross-attention instead of spatial attention features (ablation) | X-AVDT · MMDF | AUROC: 64.57 -> 91.56 | Improved |
| extract features at a later diffusion timestep (t=24 to t=499) (ablation) | X-AVDT · MMDF | AUROC: 91.56 -> 68.11 | Hurt |
| use X-AVDT instead of retrained RealForensics on MegActor-Sigma | X-AVDT · MMDF MegActor-Sigma | AUROC: 97.60 -> 90.17 | Hurt |
| use X-AVDT instead of retrained RealForensics averaged over generators | X-AVDT · MMDF average | AUROC: 92.42 -> 95.29 | Improved |
Closest papers by meaning
- AVFF: Audio-Visual Feature Fusion for Video Deepfake Detection (CVPR 2024 · similarity 0.91)
- X-AVDD: Cross-Attentive Audio-Visual Dataset Distillation (NeurIPS 2026 · similarity 0.88)
- Multi-modal Deepfake Detection via Multi-task Audio-Visual Prompt Learning (AAAI 2025 · similarity 0.88)
- Circumventing Shortcuts in Audio-visual Deepfake Detection Datasets with Unsupervised Learning (CVPR 2025 · similarity 0.87)
- Standing on the Shoulders of Giants: Reprogramming Visual-Language Model for General Deepfake Detection (AAAI 2025 · similarity 0.87)
- Adaptive Compression and Targeted Perturbation: A Unified Framework for Generalized Audio Deepfake Detection (NeurIPS 2026 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
