Research map / Video / Video segmentation and tracking
Adaptive Human Matting for Dynamic Videos
Chung-Ching Lin, Jiang Wang, Kun Luo, Kevin Lin, Linjie Li, Lijuan Wang, Zicheng Liu
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 long-term attention to short-term attention (ablation) | AdaM · VM 1920×1080 test | MAD: 5.32 -> 4.61 | Improved |
| Add short-term attention to long-term attention (ablation) | AdaM · VM 1920×1080 test | dtSSD: 1.63 -> 1.47 | Improved |
| Add the Fg/Bg mask update mechanism (ablation) | AdaM · VM 1920×1080 test | MAD: 6.15 -> 4.61 | Improved |
| Use an Fg/Bg mask rather than the alpha matte for updates (ablation) | AdaM · VM 1920×1080 test | MAD: 5.65 -> 4.61 | Improved |
| Replace the initial Mask R-CNN mask generator with Transfiner (ablation) | AdaM · VM 1920×1080 test | MAD: 4.61 -> 4.58 | No clear effect |
| Replace MobileNetV2 with ResNet50 (ablation) | AdaM · VM 1920×1080 test | MAD: 4.61 -> 4.53 | Improved |
| Add offline bidirectional inference (ablation) | AdaM · VM 1920×1080 test | MAD: 4.61 -> 4.57 | Improved |
| Add Stage 1 segmentation-data pretraining (ablation) | AdaM · VM 1920×1080 test | MAD: 4.85 -> 4.63 | Improved |
| Alternate Stage 2 training between VM and video segmentation data (ablation) | AdaM · VM 1920×1080 test | MAD: 4.63 -> 4.61 | No clear effect |
| Alternate Stage 2 training between VM and video segmentation data (ablation) | AdaM · CRGNN-R 1920×1080 | MAD: 8.07 -> 7.13 | Improved |
| Add Stage 3 HD training (ablation) | AdaM · VM 1920×1080 test | MAD: 4.61 -> 4.42 | Improved |
| Add an intermediate Fg/Bg mask output and coarse-alpha output (ablation) | AdaM · VM 1920×1080 test | MAD: 5.65 -> 4.61 | Improved |
Closest papers by meaning
- Static-to-Dynamic: Animating Still Mattes via Generative Motion for Video Matting (NeurIPS 2026 · similarity 0.91)
- Matting Anything 2: Towards Video Matting for Anything (ICLR 2026 · similarity 0.91)
- $\alpha$Matte4K & $\mu$Matting: Dataset and Model for Ultra-Micro Precision Alpha Video Matting (CVPR 2026 · similarity 0.91)
- MaGGIe: Masked Guided Gradual Human Instance Matting (CVPR 2024 · similarity 0.90)
- MatAnyone: Stable Video Matting with Consistent Memory Propagation (CVPR 2025 · similarity 0.89)
- In-Context Matting (CVPR 2024 · similarity 0.88)
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.
