Research map / Video / Video segmentation and tracking
FST-SAM3: Taming SAM~3 with Frequency-Spatio-Temporal Refinement for Video Polyp Segmentation
Guanhao Wu, Guilian Chen, Huisi Wu, Jing Qin
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 FSA frequency-spectrum adapter (ablation) | SAFE-less FST-SAM3 · SUN-SEG-Easy | Dice: 84.00 -> 86.03 | Improved |
| add ARC attention residual compensation (ablation) | SAFE-less FST-SAM3 · SUN-SEG-Hard | Dice: 82.24 -> 86.33 | Improved |
| add both FSA and ARC (ablation) | FST-SAM3 · SUN-SEG-Hard | Dice: 82.24 -> 89.39 | Improved |
| add spatial token construction (STC) (ablation) | FST-SAM3 · SUN-SEG-Easy | Dice: 86.39 -> 88.42 | Improved |
| add ungated TGU after STC (ablation) | FST-SAM3 · SUN-SEG-Hard | Dice: 86.43 -> 88.21 | Improved |
| add quality-aware gate to TGU (ablation) | FST-SAM3 · SUN-SEG-Hard | Dice: 88.21 -> 89.39 | Improved |
| apply FST-SAM3 adapters to frozen SAM 3 | SAM 3 · SUN-SEG-Hard | Dice: 75.17 -> 90.01 | Improved |
| apply FST-SAM3 adapters to frozen SAM 3 | SAM 3 · SUN-SEG-Hard | FPS: 24.59 -> 16.43 | Hurt |
| increase clip length beyond 6-7 frames (ablation) | FST-SAM3 · SUN-SEG | median Dice | Hurt |
| increase number of first-frame point prompts (ablation) | FST-SAM3 · SUN-SEG | Dice | No clear effect |
| use FST-SAM3 without any prompt versus task-specific STDDNet | FST-SAM3 · SUN-SEG-Easy | Dice: 85.86 -> 89.53 | Improved |
Closest papers by meaning
- VP-SAM: Taming Segment Anything Model for Video Polyp Segmentation via Disentanglement and Spatio-temporal Side Network (ECCV 2024 · similarity 0.91)
- One Polyp Identifies All: One-Shot Polyp Segmentation with SAM via Cascaded Priors and Iterative Prompt Evolution (ICCV 2025 · similarity 0.91)
- An Embedding-Unleashing Video Polyp Segmentation Framework via Region Linking and Scale Alignment (AAAI 2024 · similarity 0.87)
- VISTA3D: A Unified Segmentation Foundation Model For 3D Medical Imaging (CVPR 2025 · similarity 0.84)
- Boost the Inference with Co-training: A Depth-guided Mutual Learning Framework for Semi-supervised Medical Polyp Segmentation (CVPR 2025 · similarity 0.84)
- Polyper: Boundary Sensitive Polyp Segmentation (AAAI 2024 · similarity 0.83)
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.
