Research map / Video / Video segmentation and tracking
Flow-Attention-based Spatio-Temporal Aggregation Network for 3D Mask Detection
Yuxin Cao, Yian Li, Yumeng Zhu, Derui Wang, Minhui Xue
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 |
|---|---|---|---|
| remove temporal branch (spatial only) (ablation) | FASTEN · HiFiMask | HTER (%): 3.11 -> 2.11 | Improved |
| remove spatial branch (temporal only) (ablation) | FASTEN · HiFiMask | HTER (%): 7.47 -> 2.11 | Improved |
| use equal frame weights instead of flow-attention weights (ablation) | FASTEN · HiFiMask | HTER (%): 2.91 -> 2.11 | Improved |
| finetune FlowNetFace jointly instead of freezing it (ablation) | FASTEN · HiFiMask | HTER (%): 2.76 -> 2.11 | Improved |
| input frames lambda 5 vs 2 (ablation) | FASTEN · HiFiMask | HTER (%): 3.73 -> 2.11 | Improved |
| input frames lambda 5 vs 8 (ablation) | FASTEN · HiFiMask | HTER (%): 2.78 -> 2.11 | Improved |
| input frames lambda 5 vs 10 (ablation) | FASTEN · HiFiMask | HTER (%): 2.90 -> 2.11 | Improved |
| FASTEN vs HRFP (ablation) | FASTEN · HiFiMask intra-dataset | HTER (%): 2.20 -> 2.11 | Improved |
| FASTEN vs HRFP cross-dataset (ablation) | FASTEN · 3DMAD -> HKBU-MARs V1+ | HTER (%): 25.8 -> 11.8 | Improved |
| FASTEN vs HRFP cross-dataset | FASTEN · HiFiMask -> HKBU-MARs V1+ | HTER (%): 7.68 -> 7.38 but HRFP best on EER 8.33 vs 8.55 | Hurt |
| FASTEN vs HRFP inference speed | FASTEN · inference time | ms: 286.5ms -> 25.3ms | Improved |
| FASTEN vs CDCN++ inference speed | FASTEN · inference time | ms: 28.8ms -> 25.3ms | Improved |
Closest papers by meaning
- Deep Spatial Gradient and Temporal Depth Learning for Face Anti-Spoofing (CVPR 2020 · similarity 0.85)
- Learning Knowledge-Based Prompts for Robust 3D Mask Presentation Attack Detection (TPAMI 2025 · similarity 0.84)
- Spatiotemporal Graph Neural Network based Mask Reconstruction for Video Object Segmentation (AAAI 2021 · similarity 0.84)
- AltFreezing for More General Video Face Forgery Detection (CVPR 2023 · similarity 0.84)
- EyeGraph: Modularity-aware Spatio Temporal Graph Clustering for Continuous Event-based Eye Tracking (NeurIPS 2024 · similarity 0.81)
- FactorizePhys: Matrix Factorization for Multidimensional Attention in Remote Physiological Sensing (NeurIPS 2024 · similarity 0.81)
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.
