Research map / Detection and segmentation / CNN features
Efficient Mask Correction for Click-Based Interactive Image Segmentation
Fei Du, Jianlong Yuan, Zhibin Wang, Fan Wang
This paper tried a common change: Add attention module
Add the click-guided self-attention module · DAVIS · NoC@90: 6.98 -> 6.02 Improved
What 492 accepted papers report for this change
- 444 improved
- 20 no clear effect
- 4 mixed
- 24 hurt
- Multi-Scale Wavelet Transformers for Operator Learning of Dynamical Systems ICML 2026 · add wavelet attention block · CKF step 64 · Rel L2: 1.09 -> 0.35Improved
- DA-DFGAS:Differentiable Federated Graph Neural Architecture Search with Distribution-Aware Attentive Aggregation AAAI 2026 · add attention mechanism to aggregation · Table 4 datasets · parameters: about 5-10% more parametersHurt
- Caracal: Causal Architecture via Spectral Mixing ICML 2026 · add sliding window attention layers to MHF stack · Avg accuracy · avg accuracy (%): 48.22 -> 49.01Improved
See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.
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 the first click as input to the base segmentation network (ablation) | EMC-Click with HRNet18s · DAVIS | NoC@90: 7.11 -> 6.98 | Improved |
| Add the click-guided correlation module (ablation) | EMC-Click with HRNet18s · DAVIS | NoC@90: 6.98 -> 6.32 | Improved |
| Add the click-guided self-attention module (ablation) | EMC-Click with HRNet18s · DAVIS | NoC@90: 6.98 -> 6.02 | Improved |
| Add template selection to the click-guided modules (ablation) | EMC-Click with HRNet18s · DAVIS | NoC@90: 5.66 -> 5.86 | Improved |
| Combine correlation and self-attention modules (ablation) | EMC-Click with HRNet18s · DAVIS | NoC@90: 6.32 -> 5.86 | Improved |
| Add negative templates to the correlation module (ablation) | EMC-Click correlation module · SBD | NoC@90: 6.39 -> 6.26 | Improved |
| Add negative templates to the self-attention module (ablation) | EMC-Click self-attention module · Pascal VOC | NoC@90: 3.22 -> 2.99 | Improved |
| Use EMC-Click with HRNet32 instead of PseudoClick with HRNet32 | HRNet32 click-based interactive segmentation · DAVIS | NoC@90: 5.27 -> 5.33 | Hurt |
| Use EMC-Click with HRNet32 instead of PseudoClick with HRNet32 | HRNet32 click-based interactive segmentation · Pascal VOC | NoC@90: 1.94 -> 2.22 | Hurt |
Closest papers by meaning
- NTClick: Achieving Precise Interactive Segmentation With Noise-tolerant Clicks (CVPR 2025 · similarity 0.90)
- Variance-Insensitive and Target-Preserving Mask Refinement for Interactive Image Segmentation (AAAI 2024 · similarity 0.88)
- CFR-ICL: Cascade-Forward Refinement with Iterative Click Loss for Interactive Image Segmentation (AAAI 2024 · similarity 0.88)
- TETRIS: Towards Exploring the Robustness of Interactive Segmentation (AAAI 2024 · similarity 0.88)
- RClicks: Realistic Click Simulation for Benchmarking Interactive Segmentation (NeurIPS 2024 · similarity 0.87)
- FocSAM: Delving Deeply into Focused Objects in Segmenting Anything (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.
