Research map / Video / Video segmentation and tracking
Enhancing the Security of Visual Speaker Authentication Based on Dynamic Lip-Print Analysis
Yi He, Lei Yang, Bofan Chen, Shilin Wang
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 temporal frame-difference features in MD-Encoder (ablation) | Lip-print VSA · VSA unified prompts, DeepFake attacks | AUC: FaceSwap 0.9805 -> 0.9951; DeepFaceLab 0.9920 -> 0.9983 | Improved |
| add 50% static-feature dropout during training (ablation) | Lip-print VSA · VSA unified prompts, DeepFake attacks | AUC: DeepFaceLab 0.9807 -> 0.9983 | Improved |
| add modality dropout, human imposters (ablation) | Lip-print VSA · VSA unified prompts, human imposters | AUC: 0.9983 -> 0.9992 | No clear effect |
| add Subspace Learning Module (PLDA per viseme combination) (ablation) | Lip-print VSA · VSA unified prompts | AUC: human 0.9884 -> 0.9992 | Improved |
| use user-specific dynamic lip-print prompts instead of unified prompts | Lip-print VSA · VSA, human imposters | HTER (%): 1.9 -> 0.1 | Improved |
| use user-specific prompts instead of unified prompts | Siamese baseline · VSA, human imposters | HTER (%): 2.0 -> 2.5 | Mixed |
| add global branch to local branch under recombination attack (ablation) | Lip-print VSA · VSA segment recombination | AUC: 0.9626 -> 0.9835 | Improved |
| add global branch on top of local branch (combined vs global only) (ablation) | Lip-print VSA · VSA segment recombination | AUC: 0.9832 -> 0.9835 | No clear effect |
Closest papers by meaning
- Learning Forgery-Aware Lip Representations Without Forgery Priors (CVPR 2026 · similarity 0.89)
- Lips Are Lying: Spotting the Temporal Inconsistency between Audio and Visual in Lip-Syncing DeepFakes (NeurIPS 2024 · similarity 0.85)
- Restoring Speaking Lips from Occlusion for Audio-Visual Speech Recognition (AAAI 2024 · similarity 0.83)
- SpeechForensics: Audio-Visual Speech Representation Learning for Face Forgery Detection (NeurIPS 2024 · similarity 0.83)
- Personalized Lip Reading: Adapting to Your Unique Lip Movements with Vision and Language (AAAI 2025 · similarity 0.82)
- Visual Hallucination Elevates Speech Recognition (AAAI 2024 · similarity 0.82)
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.
