Research map / Detection and segmentation / CNN features
Dynamic Neural Representational Decoders for High-Resolution Semantic Segmentation
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 |
|---|---|---|---|
| replace the DeepLabV3+ decoder with NRD | ResNet-50 segmentation model · Cityscapes val | mIoU: 78.9 -> 79.8 | Helped |
| replace the DeepLabV3+ decoder with NRD | ResNet-50 segmentation model · Cityscapes val | Trimap mIoU: 49.8 -> 51.6 | Helped |
| replace bilinear decoding with NRD without guidance (ablation) | ResNet-50 segmentation model · Cityscapes val | mIoU: 74.7 -> 78.2 | Helped |
| replace CARAFE with NRD | ResNet-50 segmentation model · Cityscapes val | mIoU: 72.1 -> 75.0 | Helped |
| replace DUC with NRD | ResNet-50 segmentation model · Cityscapes val | mIoU: 74.7 -> 75.0 | Helped |
| reduce representational network channels from 16 to 8 (ablation) | NRD · Cityscapes val | mIoU: 79.8 -> 79.4 | Hurt |
| increase representational network channels from 16 to 32 (ablation) | NRD · Cityscapes val | mIoU: 79.8 -> 79.6 | Hurt |
| reduce guidance-map channels from 16 to 8 (ablation) | NRD · Cityscapes val | mIoU: 79.8 -> 79.5 | Hurt |
| increase guidance-map channels from 16 to 32 (ablation) | NRD · Cityscapes val | mIoU: 79.8 -> 79.0 | Hurt |
| add coordinate maps to an NRD using guidance maps only (ablation) | NRD · Cityscapes val | mIoU: 78.3 -> 79.8 | Helped |
| add guidance maps to an NRD using coordinate maps only (ablation) | NRD · Cityscapes val | Trimap mIoU: 46.6 -> 51.5 | Helped |
Closest papers by meaning
- Decoders Matter for Semantic Segmentation: Data-Dependent Decoding Enables Flexible Feature Aggregation (CVPR 2019 · similarity 0.90)
- Dense Decoder Shortcut Connections for Single-Pass Semantic Segmentation (CVPR 2018 · similarity 0.90)
- Rethinking Semantic Segmentation From a Sequence-to-Sequence Perspective With Transformers (CVPR 2021 · similarity 0.88)
- Rethinking Decoders for Transformer-based Semantic Segmentation: A Compression Perspective (NeurIPS 2024 · similarity 0.87)
- SED: A Simple Encoder-Decoder for Open-Vocabulary Semantic Segmentation (CVPR 2024 · similarity 0.87)
- MaSS13K: A Matting-level Semantic Segmentation Benchmark (CVPR 2025 · 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.