Research map / Detection and segmentation / CNN features
PIDNet: A Real-Time Semantic Segmentation Network Inspired by PID Controllers
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 ADB-Bag to BiSeNet(Res18) (ablation) | BiSeNet(Res18) · Cityscapes val | mIOU: 75.4 -> 76.7 | Helped |
| Add ADB-Bag to DDRNet-23 (ablation) | DDRNet-23 · Cityscapes val | mIOU: 79.5 -> 80.0 | Helped |
| Add ADB-Bag to two-branch networks (ablation) | BiSeNet(Res18), DDRNet-23 · Cityscapes val | FPS: 63.2 -> 52.1; 51.4 -> 39.2 | Hurt |
| Use Pag lateral connections instead of Add with Add fusion (ablation) | PIDNet-L · Cityscapes val | mIOU: 79.3 -> 78.1 | Hurt |
| Add ImageNet pretraining (ablation) | PIDNet-L with Add lateral and Add fusion · Cityscapes val | mIOU: 79.3 -> 80.7 | Helped |
| Use Pag lateral connections and Bag fusion instead of Add lateral and Add fusion (ablation) | PIDNet-L with ImageNet pretraining · Cityscapes val | mIOU: 80.7 -> 80.9 | Helped |
| Use DAPPM instead of PAPPM (ablation) | PIDNet-S with Bag fusion · Cityscapes val | mIOU: 78.8 -> 78.8 | No clear effect |
| Use DAPPM instead of PAPPM (ablation) | PIDNet-S with Bag fusion · Cityscapes val | FPS: 93.2 -> 83.7 | Hurt |
| Replace Bag fusion with Add fusion (ablation) | PIDNet-S with PAPPM · Cityscapes val | mIOU: 78.8 -> 78.4 | Hurt |
| Replace Bag fusion with Add fusion (ablation) | PIDNet-S with PAPPM · Cityscapes val | FPS: 93.2 -> 97.8 | Helped |
| Add auxiliary semantic loss l0 (ablation) | PIDNet-L · Cityscapes val | mIOU: 78.6 -> 78.8 | Helped |
| Add boundary loss l1 (ablation) | PIDNet-L with l0 · Cityscapes val | mIOU: 78.8 -> 79.9 | Helped |
Closest papers by meaning
- HyperSeg: Patch-Wise Hypernetwork for Real-Time Semantic Segmentation (CVPR 2021 · similarity 0.88)
- Rethinking BiSeNet for Real-Time Semantic Segmentation (CVPR 2021 · similarity 0.87)
- SCTNet: Single-Branch CNN with Transformer Semantic Information for Real-Time Segmentation (AAAI 2024 · similarity 0.87)
- Full-Resolution Residual Networks for Semantic Segmentation in Street Scenes (CVPR 2017 · similarity 0.87)
- Context-Reinforced Semantic Segmentation (CVPR 2019 · similarity 0.87)
- Efficient Piecewise Training of Deep Structured Models for Semantic Segmentation (CVPR 2016 · similarity 0.87)
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.