Research map / Detection and segmentation / Open-set and few-shot detection
SED++: A Simple Encoder-Decoder for Improved Open-Vocabulary Semantic Segmentation
Wenqi Zhu, Bin Xie, Jiale Cao, Jin Xie, Fahad Shahbaz Khan, Yanwei Pang
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 |
|---|---|---|---|
| replace plain ViT with hierarchical ConvNeXt encoder (ablation) | SED baseline · A-150 | mIoU: 23.7 -> 28.2 | Improved |
| add gradual fusion decoder (ablation) | SED with hierarchical encoder · A-150 | mIoU: 28.2 -> 31.8 | Improved |
| add category early rejection (ablation) | SED · A-150 | mIoU: 31.8 -> 31.6 | No clear effect |
| add skip-layer fusion to hierarchical encoder (ablation) | SED · A-150 | mIoU: 25.7 -> 28.2 | Improved |
| add skip-layer fusion to plain ViT encoder (ablation) | plain ViT cost map · A-150 | mIoU: 23.0 -> 23.7 | Improved |
| fine-tune whole encoder instead of freezing it (ablation) | SED · A-150 | mIoU: 28.6 -> 31.8 | Improved |
| encoder learning rate factor 0.01 instead of 0.005 or 0.02 (ablation) | SED · PC-459 | mIoU: 17.6 -> 18.6 | Improved |
| add class-level aggregation to spatial-level (ablation) | FAM · A-150 | mIoU: 28.8 -> 31.8 | Improved |
| add spatial-level aggregation to class-level (ablation) | FAM · A-150 | mIoU: 30.7 -> 31.8 | Improved |
| stop gradient from skip-layer fusion to encoder (ablation) | SFM · PC-459 | mIoU: 18.0 -> 18.6 | Improved |
| use three decoder layers instead of one (ablation) | GFD · PC-459 | mIoU: 17.0 -> 18.6 | Improved |
| category early rejection with top-k=8 instead of all categories (ablation) | SED · A-847 | inference time (ms): 861.0 -> 181.6 | Improved |
Closest papers by meaning
- SED: A Simple Encoder-Decoder for Open-Vocabulary Semantic Segmentation (CVPR 2024 · similarity 0.99)
- Exploring Simple Open-Vocabulary Semantic Segmentation (CVPR 2025 · similarity 0.91)
- Open-Vocabulary Semantic Segmentation with Image Embedding Balancing (CVPR 2024 · similarity 0.90)
- CAT-Seg: Cost Aggregation for Open-Vocabulary Semantic Segmentation (CVPR 2024 · similarity 0.90)
- Effective SAM Combination for Open-Vocabulary Semantic Segmentation (CVPR 2025 · similarity 0.90)
- Training-free Open-Vocabulary Semantic Segmentation via Diverse Prototype Construction and Sub-region Matching (AAAI 2025 · 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.
