Research map / Detection and segmentation / CNN features
Interactive Segmentation As Gaussion Process Classification
Minghao Zhou, Hong Wang, Qian Zhao, Yuexiang Li, Yawen Huang, Deyu Meng, Yefeng Zheng
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 |
|---|---|---|---|
| Use GPCIS instead of RITM | interactive segmentation with ResNet50 · DAVIS | NoC@90: 6.05 -> 5.89 | Improved |
| Use GPCIS instead of RITM | interactive segmentation with ResNet50 · Berkeley | NoC@90: 2.95 -> 2.60 | Improved |
| Use GPCIS instead of FocusCut | interactive segmentation with ResNet50 · SBD | NoC@85: 3.62 -> 3.80 | Hurt |
| Use GPCIS instead of FocusCut | interactive segmentation with ResNet50 · GrabCut | NoC@90: 1.78 -> 1.82 | Hurt |
| Use GPCIS instead of RITM | interactive segmentation with ResNet50 · DAVIS | IoU&1: 72.89% -> 75.67% | Improved |
| Use GPCIS instead of RITM | interactive segmentation with ResNet50 · DAVIS | SPC: 34.24 -> 38.82 | Hurt |
| Add LVI guidance (ablation) | GPCIS · GrabCut, Berkeley, SBD, DAVIS (average) | NoC@85: 2.98 -> 2.85 | Improved |
| Add LVI guidance (ablation) | GPCIS · GrabCut, Berkeley, SBD, DAVIS (average) | NoC@90: 4.07 -> 4.00 | Improved |
| Learn function-space kernel parameters with DKL-F (ablation) | GPCIS · GrabCut, Berkeley, SBD, DAVIS (average) | NoC@90: 4.16 -> 4.00 | Improved |
| Learn weight-space prior parameters with DKL-W (ablation) | GPCIS · GrabCut, Berkeley, SBD, DAVIS (average) | NoC@90: 4.34 -> 4.00 | Improved |
| Concatenate input RGB with deep features (ablation) | GPCIS · GrabCut, Berkeley, SBD, DAVIS (average) | NoC@90: 4.10 -> 4.00 | Improved |
| Reduce testing epsilon squared (ablation) | GPCIS with ResNet50 · DAVIS | NoIC: 36 -> 30 -> 21 -> 15 ->15 -> 8 -> 2 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- FocusCut — FocusCut: Diving Into a Focus View in Interactive Segmentation (CVPR 2022) · better on 1 of 5 shared results
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
- TETRIS: Towards Exploring the Robustness of Interactive Segmentation (AAAI 2024 · similarity 0.87)
- NTClick: Achieving Precise Interactive Segmentation With Noise-tolerant Clicks (CVPR 2025 · similarity 0.86)
- RClicks: Realistic Click Simulation for Benchmarking Interactive Segmentation (NeurIPS 2024 · similarity 0.86)
- CFR-ICL: Cascade-Forward Refinement with Iterative Click Loss for Interactive Image Segmentation (AAAI 2024 · similarity 0.85)
- GraCo: Granularity-Controllable Interactive Segmentation (CVPR 2024 · similarity 0.85)
- Repurposing Stable Diffusion Attention for Training-Free Unsupervised Interactive Segmentation (CVPR 2025 · 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.
