Research map / Image generation / Editing and style transfer
Infusing Definiteness into Randomness: Rethinking Composition Styles for Deep Image Matting
Zixuan Ye, Yutong Dai, Chaoyi Hong, Zhiguo Cao, Hao Lu
This paper tried a common change: Add data augmentation
Add data augmentation to the unprocessed baseline · Composition-1K · SAD: 44.87 -> 43.03 Improved
What 689 accepted papers report for this change
- 560 improved
- 44 no clear effect
- 26 mixed
- 59 hurt
- Understanding Multimodal Learning: A Loss Landscape Smoothness Perspective ICML 2026 · add data augmentation (jitter, crop, flip) · UPMC-Food101 · accuracy: 92.84 -> 94.93Improved
- AsyncBEV: Cross-modal flow alignment in Asynchronous 3D Object Detection ICLR 2026 · add EMC + asynchronous data augmentation training · nuScenes val, 0 s offset · NDS: 66.7 -> 63.1Hurt
- Towards Understanding Continual Factual Knowledge Acquisition of Language Models: From Theory to Algorithm ICML 2026 · add data augmentation (5 biographies per individual) in pre-training · Biography PT test, unseen templates · EM: 2.71 -> 81.27Improved
See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.
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 |
|---|---|---|---|
| Use triplet-style foreground selection instead of DIM-style | IndexNet Matting · Composition-1K | SAD: 43.03 -> 38.65 | Improved |
| Use quadruplet-style foreground selection instead of GCA-style | MatteFormer · Composition-1K | SAD: 23.71 -> 22.60 | Improved |
| Use triplet-style composition instead of GCA-style | MatteFormer · AIM-500 | SAD: 28.39 -> 26.33 | Improved |
| Use triplet-style composition instead of GCA-style | MatteFormer · PPM-100 | SAD: 41.01 -> 38.94 | Improved |
| Replace naive foreground combination (NCF) with reasonable foreground combination (RCF) | A2U with quadruplet-style · Composition-1K | SAD: 32.70 -> 30.47 | Improved |
| Replace naive foreground combination (NCF) with reasonable foreground combination (RCF) | IndexNet with GCA-style · Composition-1K | SAD: 45.19 -> 44.03 | Improved |
| Add data augmentation to the unprocessed baseline (ablation) | IndexNet · Composition-1K | SAD: 44.87 -> 43.03 | Improved |
| Add RCF and triplet-style composition (ablation) | IndexNet with data augmentation · Composition-1K | SAD: 43.03 -> 38.65 | Improved |
| Replace triplet-style with quadruplet-style composition (ablation) | IndexNet with data augmentation and RCF · Composition-1K | SAD: 38.65 -> 38.16 | Improved |
| Order related samples rather than shuffle them (ablation) | A2U with triplet-style composition · Composition-1K | SAD: 31.52 -> 32.40 | Hurt |
| Order related samples rather than shuffle them (ablation) | A2U with quadruplet-style composition · Composition-1K | SAD: 30.47 -> 31.56 | Hurt |
| Replace triplet-style with quadruplet-style composition (ablation) | MatteFormer · PPM-100 | SAD: 38.94 -> 40.22 | Hurt |
Closest papers by meaning
- Image Matting without Matting-Specific Annotations via Eikonal Fields (NeurIPS 2026 · similarity 0.88)
- Training Matting Models Without Alpha Labels (AAAI 2025 · similarity 0.88)
- Privileged Prior Information Distillation for Image Matting (AAAI 2024 · similarity 0.87)
- DRIP: Unleashing Diffusion Priors for Joint Foreground and Alpha Prediction in Image Matting (NeurIPS 2024 · similarity 0.86)
- Revisiting Context Aggregation for Image Matting (ICML 2024 · similarity 0.86)
- MaGGIe: Masked Guided Gradual Human Instance Matting (CVPR 2024 · similarity 0.86)
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.
