Research map / Detection and segmentation / CNN features
Spatial Frequency Modulation for Semantic Segmentation
Linwei Chen, Ying Fu, Lin Gu, Dezhi Zheng, Jifeng Dai
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 SFM | InternImage-T · ADE20K val | mIoU: 47.9 -> 49.2 (+1.3) | Improved |
| add SFM | InternImage-L · ADE20K val | mIoU: 53.9 -> 54.7 | Improved |
| add SFM | Mask2Former Swin-L · ADE20K val | mIoU: 56.1 -> 56.7 | Improved |
| add SFM | Mask2Former Swin-T · ADE20K val | mIoU: 47.7 -> 49.1 (+1.4) | Improved |
| replace Blur/AdaBlur/FLC low-pass downsampling with SFM | UPerNet-ResNet50 · Cityscapes val | mIoU: 79.5 vs 78.8, 78.9, 78.6 | Improved |
| add ARS to strided-conv downsampling vs DeformConv | PCAA-R50 stride 32 · Cityscapes val | mIoU: 73.6 -> 74.7 | Improved |
| add ARS to DeformConv | PCAA-R50 stride 8 · Cityscapes val | mIoU: 79.5 -> 80.3 | Improved |
| add SFM | FCN stride 32 · Cityscapes val | mIoU: 71.2 -> 76.2 (+5.0) | Improved |
| add SFM, stride 8 vs stride 32 | PSPNet · Cityscapes val | mIoU: +3.3 at stride 32 (72.7 -> 76.0), +1.1 at stride 8 (78.4 -> 79.5) | Mixed |
| add ARS | ResNet-50 · CIFAR-100 | top-1 accuracy: 81.36 -> 82.18 | Improved |
| add ARS | ResNeXt-50 · Tiny-ImageNet | top-1 accuracy: 75.39 -> 75.79 | Improved |
| add ARS under PGD attack | Preact-ResNet-18 FGSM-AT · CIFAR-10 | PGD robust accuracy: 0.02 -> 32.74 | Improved |
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.
- Mask DINO — Mask DINO: Towards a Unified Transformer-Based Framework for Object Detection and Segmentation (CVPR 2023) · better on 1 of 1 shared result
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
- When Semantic Segmentation Meets Frequency Aliasing (ICLR 2024 · similarity 0.89)
- Frequency-Adaptive Dilated Convolution for Semantic Segmentation (CVPR 2024 · similarity 0.88)
- FFR: Frequency Feature Rectification for Weakly Supervised Semantic Segmentation (CVPR 2025 · similarity 0.86)
- Learning Frequency-Adapted Vision Foundation Model for Domain Generalized Semantic Segmentation (NeurIPS 2024 · similarity 0.86)
- Beyond Spatial Domain: Cross-domain Promoted Fourier Convolution Helps Single Image Dehazing (AAAI 2025 · similarity 0.85)
- Lightweight Frequency Masker for Cross-Domain Few-Shot Semantic Segmentation (NeurIPS 2024 · 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.
