Research map / Detection and segmentation / Open-set and few-shot detection
Personalized Image Semantic Segmentation
Yu Zhang, Chang-Bin Zhang, Peng-Tao Jiang, Ming-Ming Cheng, Feng Mao
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 group region context module (ablation) | ADVENT baseline (VGG-16) · PSS | FIoU: 44.22 -> 47.16 | Improved |
| use global group context instead of no context (ablation) | ADVENT baseline (VGG-16) · PSS | FIoU: 44.22 -> 44.56 | No clear effect |
| add pseudo label refinement step (ablation) | OURS-S1 (ResNet-50) · PSS | FIoU: 58.79 -> 59.72 | Improved |
| add pseudo label refinement step (ablation) | OURS-S1 (ResNet-50) · PSS | MIoU: 37.46 -> 39.16 | Improved |
| add group context module to ADVENT (ablation) | ADVENT (ResNet-50) · PSS | MIoU: 37.31 -> 37.46 | No clear effect |
| add group context module to ADVENT (ablation) | ADVENT (ResNet-50) · PSS | FIoU: 56.59 -> 58.79 | Improved |
| train on personal data instead of mixing all users' images (ablation) | OURS (VGG-16) · PSS | FIoU: 45.56 -> 47.16 | Improved |
| use 80 groups instead of 1 group (ablation) | OURS (VGG-16) · PSS | FIoU: 46.64 -> 47.16 | Improved |
| use 80 groups instead of 200 groups (ablation) | OURS (VGG-16) · PSS | FIoU: 46.66 -> 47.16 | Improved |
| add group context module to ADVENT (ablation) | ADVENT (ResNet-101) · GTA5 to Cityscapes | mIoU: 43.8 -> 45.4 | Improved |
| add group context module (OURS-S1 vs baseline without context) (ablation) | Baseline (ResNet-50), user ID6 · PSS ID6 | MIoU: 40.52 -> 33.64 | Hurt |
| add group context module (OURS-S1 vs baseline without context) (ablation) | Baseline (ResNet-50), user ID6 · PSS ID6 | FIoU: 55.04 -> 60.38 | 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.
- ADVENT — ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation (CVPR 2019) · better on 5 of 5 shared results
- MRNet — MR-NET: Exploiting Mutual Relation for Visual Relationship Detection (AAAI 2019) · 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
- Rethinking Self-Supervised Semantic Segmentation: Achieving End-to-End Segmentation (TPAMI 2024 · similarity 0.87)
- Personalize Segment Anything Model with One Shot (ICLR 2024 · similarity 0.86)
- FoundYou: A Unified Model for Personalized Segmentation and Retrieval (ECCV 2026 · similarity 0.86)
- Where's Waldo: Diffusion Features For Personalized Segmentation and Retrieval (NeurIPS 2024 · similarity 0.85)
- Label-Efficient Few-Shot Semantic Segmentation with Unsupervised Meta-Training (AAAI 2024 · similarity 0.84)
- Hierarchy-Agnostic Unsupervised Segmentation: Parsing Semantic Image Structure (NeurIPS 2024 · 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.
