Research map / Detection and segmentation / Open-set and few-shot detection
PseudoClick: Interactive Image Segmentation with Click Imitation
Qin Liu, Meng Zheng, Benjamin Planche, Srikrishna Karanam, Terrence Chen, Marc Niethammer, Ziyan Wu
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 pseudo clicks (C+L HRNet32) (ablation) | PseudoClick · DAVIS | NoC@85: 4.09 -> 3.79 | Improved |
| add pseudo clicks (C+L HRNet32) (ablation) | PseudoClick · DAVIS | NoC@90: 5.27 -> 5.11 | Improved |
| add pseudo clicks (C+L HRNet32) (ablation) | PseudoClick · Pascal | NoC@90: 2.52 -> 2.25 | Improved |
| add one pseudo click instead of FP/FN post-processing (ablation) | PseudoClick · BraTS (3 human clicks) | mIoU: 63.5 -> 64.0 | Improved |
| add FP/FN post-processing to baseline (ablation) | PseudoClick · BraTS (2 human clicks) | mIoU: 23.2 -> 42.6 | Improved |
| add one pseudo click to baseline (ablation) | PseudoClick · BraTS (2 human clicks) | mIoU: 23.2 -> 44.8 | Improved |
| train with normalized focal loss instead of BCE (ablation) | PseudoClick · DAVIS | NoC@85: 3.97 -> 3.79 | Improved |
| train with Soft IoU loss instead of focal loss (ablation) | PseudoClick · DAVIS | NoC@85: 3.79 -> 3.96 | Hurt |
| train on COCO+LVIS instead of SBD (ablation) | PseudoClick · DAVIS | NoC@85: 4.74 -> 3.79 | Improved |
| train on COCO+LVIS instead of Pascal (ablation) | PseudoClick · Pascal | NoC@85: 2.66 -> 1.94 | Improved |
| use SegFormer instead of ResNet101 backbone (ablation) | PseudoClick · Berkeley | NoC@85: 1.85 -> 2.54 | Hurt |
| use HRFormer instead of ResNet101 backbone (ablation) | PseudoClick · DAVIS | NoC@85: 4.14 -> 4.80 | No clear effect |
Closest papers by meaning
- NTClick: Achieving Precise Interactive Segmentation With Noise-tolerant Clicks (CVPR 2025 · similarity 0.93)
- RClicks: Realistic Click Simulation for Benchmarking Interactive Segmentation (NeurIPS 2024 · similarity 0.92)
- TETRIS: Towards Exploring the Robustness of Interactive Segmentation (AAAI 2024 · similarity 0.90)
- AGILE3D: Attention Guided Interactive Multi-object 3D Segmentation (ICLR 2024 · similarity 0.88)
- Order-aware Interactive Segmentation (ICLR 2025 · similarity 0.88)
- CFR-ICL: Cascade-Forward Refinement with Iterative Click Loss for Interactive Image Segmentation (AAAI 2024 · 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.
