Research map / Detection and segmentation / CNN features
CTNet: Context-Based Tandem Network for Semantic Segmentation
Zechao Li, Yanpeng Sun, Liyan Zhang, Jinhui Tang
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 SCM after CCM (tandem) instead of CCM only (ablation) | CTNet (ResNet-101) · PASCAL-Context | mIoU (MS): 55.6 -> 56.2 | Improved |
| add CCM before SCM instead of SCM only (ablation) | CTNet (ResNet-101) · PASCAL-Context | mIoU (MS): 54.5 -> 56.2 | Improved |
| add SCM to CCM (ablation) | CTNet (ResNet-50) · PASCAL-Context | mIoU (MS): 52.6 -> 53.3 | Improved |
| add CCM to SCM (ablation) | CTNet (ResNet-50) · PASCAL-Context | mIoU (MS): 52.3 -> 53.3 | Improved |
| connect CCM and SCM in tandem rather than in parallel (ablation) | CTNet (ResNet-50) · PASCAL-Context | mIoU (MS): 52.6 -> 53.3 | Improved |
| connect CCM and SCM in tandem rather than in parallel (ablation) | CTNet (ResNet-101) · PASCAL-Context | mIoU (MS): 55.4 -> 56.2 | Improved |
| use multi-scale MCE block instead of single-scale ECA block (ablation) | CTNet (ResNet-50) · PASCAL-Context | mIoU: 52.2 -> 53.3 | Improved |
| use local channel context (ECA) instead of SE global context (ablation) | CTNet (ResNet-50) · PASCAL-Context | mIoU: 51.8 -> 52.2 | Improved |
| fuse five kernel scales instead of four in MCE (ablation) | CTNet MCE · PASCAL-Context | mIoU: 53.3 -> 52.4 | Hurt |
| use global average pooling instead of max pooling in CCM (ablation) | CTNet (ResNet-50) · PASCAL-Context | mIoU: 52.3 -> 53.3 | Improved |
| use global average pooling instead of mix pooling in CCM (ablation) | CTNet (ResNet-50) · PASCAL-Context | mIoU: 52.5 -> 53.3 | Improved |
| add CP-Loss (beta=0.1) to the loss (ablation) | CTNet · PASCAL-Context | mIoU | Improved |
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.
- CCNet — CCNet: Criss-Cross Attention for Semantic Segmentation (ICCV 2019) · better on 1 of 1 shared result
- DANet — DANet: Divergent Activation for Weakly Supervised Object Localization (ICCV 2019) · 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
- Context-Reinforced Semantic Segmentation (CVPR 2019 · similarity 0.89)
- Context Encoding for Semantic Segmentation (CVPR 2018 · similarity 0.89)
- TAGNet: Learning Configurable Context Pathways for Semantic Segmentation (TPAMI 2022 · similarity 0.88)
- Learning to Predict Context-adaptive Convolution for Semantic Segmentation (ECCV 2020 · similarity 0.88)
- Multi-Scale Context Intertwining for Semantic Segmentation (ECCV 2018 · similarity 0.88)
- Contextrast: Contextual Contrastive Learning for Semantic Segmentation (CVPR 2024 · 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.
