Research map / Label-efficient and robust learning / Recognition and long-tail
Frequency-Aware Discriminative Feature Learning Supervised by Single-Center Loss for Face Forgery Detection
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 SCL to softmax loss (ablation) | Xception on FF++ c40 · FF++ c40 | AUC: 0.860 -> 0.916 | Helped |
| Add SCL to softmax loss (ablation) | Xception on FF++ c40 · FF++ c40 | pAUC_0.1: 0.652 -> 0.790 | Helped |
| Use center loss instead of SCL with softmax loss (ablation) | Xception on FF++ c40 · FF++ c40 | AUC: 0.860 -> 0.868 | Helped |
| Use triplet loss instead of SCL with softmax loss (ablation) | Xception on FF++ c40 · FF++ c40 | AUC: 0.860 -> 0.863 | No clear effect |
| Add AFFGM to the baseline (ablation) | Xception on FF++ c40 · FF++ c40 | AUC: 0.861 -> 0.906 | Helped |
| Add SCL to the baseline (ablation) | Xception on FF++ c40 · FF++ c40 | AUC: 0.861 -> 0.916 | Helped |
| Combine SCL and AFFGM (ablation) | Xception on FF++ c40 · FF++ c40 | AUC: 0.861 -> 0.924 | Helped |
| Use a 1x1 convolution fusion module with group=1 (ablation) | FDFL under softmax supervision · FF++ c40 | AUC: 0.892 -> 0.906 | Helped |
| Vary lambda from 0.001 to 1 (ablation) | SCL + softmax on Xception · FF++ c40 | AUC | No clear effect |
| Use SCL with m=0.3 and lambda=0.5 (ablation) | Xception · FF++ c40 | AUC: 0.916 | Helped |
| Train with softmax loss only, omitting SCL (ablation) | FDFL objective · FF++ c40 | AUC: 0.916 -> 0.861 | Hurt |
Closest papers by meaning
- Stacking Brick by Brick: Aligned Feature Isolation for Incremental Face Forgery Detection (CVPR 2025 · similarity 0.87)
- Identity-Aware Vision-Language Model for Explainable Face Forgery Detection (AAAI 2026 · similarity 0.86)
- Critical Forgetting-Based Multi-Scale Disentanglement for Deepfake Detection (AAAI 2025 · similarity 0.85)
- Learning Discriminative Noise Guidance for Image Forgery Detection and Localization (AAAI 2024 · similarity 0.85)
- DevFD : Developmental Face Forgery Detection by Learning Shared and Orthogonal LoRA Subspaces (NeurIPS 2025 · similarity 0.85)
- Face Forgery Video Detection via Temporal Forgery Cue Unraveling (CVPR 2025 · similarity 0.85)
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.