Research map / Detection and segmentation / CNN features
ACFNet: Attentional Class Feature Network for Semantic Segmentation
Fan Zhang, Yanqin Chen, Zhihang Li, Zhibin Hong, Jingtuo Liu, Feifei Ma, Junyu Han, Errui Ding
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 class center concatenated with features (ablation) | ResNet-101 · Cityscapes val | mIoU: 75.85 -> 77.94 | Improved |
| add attentional class feature (sum) on top of class center (ablation) | ResNet-101 · Cityscapes val | mIoU: 77.94 -> 79.32 | Improved |
| use ACF(sum) instead of ACF(concat) (ablation) | ResNet-101 + ACF · Cityscapes val | mIoU: 79.17 -> 79.32 | No clear effect |
| add ACF module to ResNet-101 (fine output) | ResNet-101 · Cityscapes val | mIoU: 75.85 -> 79.32 | Improved |
| refine coarse to fine segmentation with ACF | ResNet-101 + ACF (sum) · Cityscapes val | mIoU: 76.56 -> 79.32 | Improved |
| add ACF module after ASPP | ResNet-101 + ASPP · Cityscapes val | mIoU: 78.42 -> 80.08 | Improved |
| add online bootstrapping | ResNet-101 + ASPP + ACF · Cityscapes val | mIoU: 80.08 -> 80.99 | Improved |
| add multi-scale and flip testing | ResNet-101 + ASPP + ACF · Cityscapes val | mIoU: 80.08 -> 81.46 | Improved |
| compare ACFNet with PSANet under train-fine only training | ACFNet · Cityscapes test | mIoU: 78.6 -> 80.8 | Improved |
| compare ACFNet with DenseASPP using train-fine and val-fine | ACFNet · Cityscapes test | mIoU: 80.6 -> 81.8 | Improved |
| compare ACFNet with DenseASPP on truck class | ACFNet · Cityscapes test | truck IoU: 78.0 -> 76.6 | Hurt |
Later papers that compared with this method
4 later accepted papers report ACFNet as a baseline in their results tables; 4 of them report a better number on most of the results they share with it.
- Exploit Visual Dependency Relations for Semantic Segmentation (CVPR 2021 · better on 1 of 1 shared result)
- Cars Can't Fly Up in the Sky: Improving Urban-Scene Segmentation via Height-Driven Attention Networks (CVPR 2020 · better on 1 of 1 shared result)
- Disentangled Non-local Neural Networks (ECCV 2020 · better on 1 of 1 shared result)
- Tensor Low-Rank Reconstruction for Semantic Segmentation (ECCV 2020 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- PSANet — PSANet: Point-wise Spatial Attention Network for Scene Parsing (ECCV 2018) · better on 2 of 2 shared results
- BiSeNet — BiSeNet: Bilateral Segmentation Network for Real-time Semantic Segmentation (ECCV 2018) · 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
- E-CRF: Embedded Conditional Random Field for Boundary-caused Class Weights Confusion in Semantic Segmentation (ICLR 2023 · similarity 0.88)
- Partial Class Activation Attention for Semantic Segmentation (CVPR 2022 · similarity 0.88)
- L2G: A Simple Local-to-Global Knowledge Transfer Framework for Weakly Supervised Semantic Segmentation (CVPR 2022 · similarity 0.88)
- Fully Attentional Network for Semantic Segmentation (AAAI 2022 · similarity 0.88)
- CAR: Class-Aware Regularizations for Semantic Segmentation (ECCV 2022 · similarity 0.88)
- Squeeze-and-Attention Networks for Semantic Segmentation (CVPR 2020 · similarity 0.88)
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.
