Research map / Image generation / Image restoration
Tripartite Information Mining and Integration for Image Matting
Yuhao Liu, Jiake Xie, Xiao Shi, Yu Qiao, Yujie Huang, Yong Tang, Xin Yang
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 RGB-Unit (ablation) | Basic U-Net · Composition-1K test | SAD: 38.60 -> 33.12 | Improved |
| add Trimap-Unit (ablation) | Basic U-Net · Composition-1K test | SAD: 38.60 -> 36.77 | Improved |
| add both units with additive fusion (ablation) | Basic U-Net · Composition-1K test | SAD: 38.60 -> 31.21 | Improved |
| replace additive fusion with TI2 in the encoder (ablation) | Basic + additive integration · Composition-1K test | SAD: 31.21 -> 30.18 | Improved |
| replace additive fusion with TI2 in the decoder (ablation) | Basic + additive integration · Composition-1K test | SAD: 31.21 -> 30.60 | Improved |
| apply TI2 in both encoder and decoder (ablation) | TI2 in encoder only · Composition-1K test | SAD: 30.18 -> 29.08 | Improved |
| add RGB-Unit (ablation) | Basic U-Net · Distinctions-646 test | SAD: 32.20 -> 26.93 | Improved |
| add Trimap-Unit (ablation) | Basic U-Net · Distinctions-646 test | SAD: 32.20 -> 27.67 | Improved |
| apply TI2 in both encoder and decoder (ablation) | TI2 in encoder only · Distinctions-646 test | SAD: 25.27 -> 22.28 | Improved |
| add RGB-Unit (ablation) | Basic U-Net · Human-2K test | SAD: 5.87 -> 5.16 | Improved |
| apply TI2 in both encoder and decoder (ablation) | TI2 in encoder only · Human-2K test | SAD: 4.65 -> 4.20 | Improved |
| train on Human-2K instead of Composition-1K | DIM · Composition-1K human test images | SAD: 15.25 -> 11.46 | Improved |
Closest papers by meaning
- Bridging Composite and Real: Towards End-to-End Deep Image Matting (IJCV 2022 · similarity 0.90)
- Attention-Guided Hierarchical Structure Aggregation for Image Matting (CVPR 2020 · similarity 0.89)
- Deep Image Matting (CVPR 2017 · similarity 0.89)
- Revisiting Context Aggregation for Image Matting (ICML 2024 · similarity 0.89)
- Training Matting Models Without Alpha Labels (AAAI 2025 · similarity 0.89)
- Boosting Semantic Human Matting With Coarse Annotations (CVPR 2020 · similarity 0.89)
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.
