Research map / Detection and segmentation / CNN features
Ultra-High Resolution Image Segmentation via Locality-Aware Context Fusion and Alternating Local Enhancement
Wenxi Liu, Qi Li, Xindai Lin, Weixiang Yang, Shengfeng He, Yuanlong Yu
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 local, medium and large contexts (ablation) | FCN-8s baseline · DeepGlobe | mIoU: 71.84 -> 73.47 | Improved |
| add local, medium and large contexts (ablation) | FCN-8s baseline · Inria Aerial | mIoU: 69.08 -> 73.50 | Improved |
| add gated positional embeddings (GPE) (ablation) | three-context LCF · DeepGlobe | mIoU: 73.22 -> 73.47 | Improved |
| add gated positional embeddings (GPE) (ablation) | three-context LCF · Inria Aerial | mIoU: 73.53 -> 73.50 | No clear effect |
| use medium or large context alone instead of local context (ablation) | LCF · Inria Aerial | mIoU: 72.50 -> 72.48 | Hurt |
| use adaptive fusion instead of averaging (ablation) | LCF · DeepGlobe | mIoU: 72.89 -> 73.47 | Improved |
| use adaptive fusion instead of averaging (ablation) | LCF · Inria Aerial | mIoU: 73.20 -> 73.50 | Improved |
| use adaptive fusion instead of offline weighted fusion (ablation) | LCF · DeepGlobe | mIoU: 73.16 -> 73.47 | Improved |
| replace ALE with feature concatenation (ablation) | LCF · DeepGlobe | mIoU: 73.47 -> 72.20 | Hurt |
| add ALE to LCF (variant d) (ablation) | LCF · DeepGlobe | mIoU: 73.47 -> 73.91 | Improved |
| add ALE to LCF (variant d) (ablation) | LCF · Inria Aerial | mIoU: 73.50 -> 74.58 | Improved |
| use third-order instead of second-order ALE (ablation) | LCF + ALE · DeepGlobe | mIoU: 73.91 -> 73.77 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Segformer — SegFormer: A Topic Segmentation Model with Controllable Range of Attention (AAAI 2023) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Ultra-High Resolution Segmentation via Boundary-Enhanced Patch-Merging Transformer (AAAI 2025 · similarity 0.90)
- Toward Real Ultra Image Segmentation: Leveraging Surrounding Context to Cultivate General Segmentation Model (NeurIPS 2024 · similarity 0.88)
- Boosting the Dual-Stream Architecture in Ultra-High Resolution Segmentation with Resolution-Biased Uncertainty Estimation (CVPR 2025 · similarity 0.86)
- Hybrid Global-Local Representation with Augmented Spatial Guidance for Zero-Shot Referring Image Segmentation (CVPR 2025 · similarity 0.86)
- Hierarchical Histogram Threshold Segmentation - Auto-terminating High-detail Oversegmentation (CVPR 2024 · similarity 0.86)
- Native Segmentation Vision Transformers (NeurIPS 2025 · similarity 0.85)
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.
