Research map / Detection and segmentation / Open-set and few-shot detection
PaintSeg: Painting Pixels for Training-free Segmentation
Xiang Li, Chung-Ching Lin, Yinpeng Chen, Zicheng Liu, Jinglu Wang, Rita Singh, Bhiksha Raj
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 adversarial constraint on mask updating (vs I+O steps alone) (ablation) | AMCP · ECSSD coarse mask | IoU: 79.5 -> 80.6 | Improved |
| use both alternating I-step and O-step instead of one step only (ablation) | AMCP · ECSSD coarse mask | IoU: 78.4 -> 79.5 | Improved |
| use a single step only (second variant) (ablation) | AMCP · ECSSD coarse mask | IoU: 79.5 -> 77.9 | Hurt |
| cluster centers K=2 vs 3 (ablation) | AMCP · ECSSD coarse mask | IoU: 80.6 -> 72.3 when K=3 | Hurt |
| cluster centers K=4 (ablation) | AMCP · ECSSD coarse mask | IoU: 80.6 -> 61.5 | Hurt |
| step number T from 3 to 5 (ablation) | AMCP · ECSSD coarse mask | IoU: 78.4 -> 80.6 | Improved |
| step number T 5 to 6 (ablation) | AMCP · ECSSD coarse mask | IoU: 80.6 -> 80.5 | No clear effect |
| diffusion iterations 10 to 50 (ablation) | generative painting · ECSSD coarse mask | IoU: 77.3 -> 80.6 | Improved |
| diffusion iterations 30 to 50 (ablation) | generative painting · ECSSD coarse mask | IoU: 78.7 -> 80.6 | Improved |
| contrast box size rate 1.1 vs tight 0.9 (ablation) | AMCP · ECSSD coarse mask | IoU: 69.2 -> 80.6 | Improved |
| contrast box size rate 1.1 vs 1.0 (ablation) | AMCP · ECSSD coarse mask | IoU: 72.4 -> 80.6 | Improved |
| contrast box size rate 1.2 vs 1.1 (ablation) | AMCP · ECSSD coarse mask | IoU: 80.6 -> 80.0 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- TokenCut — TokenCut: Segmenting Objects in Images and Videos With Self-Supervised Transformer and Normalized Cut (TPAMI 2023) · better on 4 of 4 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
- Paint by Inpaint: Learning to Add Image Objects by Removing Them First (CVPR 2025 · similarity 0.86)
- Object Segmentation Without Labels with Large-Scale Generative Models (ICML 2021 · similarity 0.85)
- Segment Anything without Supervision (NeurIPS 2024 · similarity 0.84)
- Integrating Low-Level Visual Cues for Enhanced Unsupervised Semantic Segmentation (AAAI 2025 · similarity 0.84)
- Hierarchy-Agnostic Unsupervised Segmentation: Parsing Semantic Image Structure (NeurIPS 2024 · similarity 0.83)
- Prompting Diffusion Models for Zero-Shot Instance Segmentation (NeurIPS 2026 · similarity 0.83)
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.
