Research map / Detection and segmentation / Open-set and few-shot detection
FLOAT: Factorized Learning of Object Attributes for Improved Multi-Object Multi-Part Scene Parsing
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Factorize object and part labels into object, shared part-root, and side-attribute predictions (ablation) | Multi-object multi-part parsing · Pascal-Part-201 | mIOU: 26.3 -> 37.1 | Helped |
| Add inference-time zoom refinement (ablation) | FLOAT · Pascal-Part-201 | sqIOU: 27.8 -> 29.6 | Helped |
| Add inference-time zoom refinement (ablation) | FLOAT · Pascal-Part-108 | sqIOU: 38.4 -> 40.5 | Helped |
| Add inference-time zoom refinement (ablation) | FLOAT · Pascal-Part-58 | sqIOU: 51.7 -> 54.2 | Helped |
| Add a side-attribute decoder (ablation) | FLOAT · Pascal-Part-201 | mIOU: 31.3 -> 36.9 | Helped |
| Replace IZR with Random Crop Zoom (ablation) | FLOAT · Pascal-Part-201 | mIOU: 36.6 -> 37.1 | Hurt |
| Replace IZR with Random Crop Zoom (ablation) | FLOAT · Pascal-Part-201 | sqIOU: 28.0 -> 29.6 | Mixed |
| Use separate left-right and front-back side decoders (ablation) | FLOAT · Pascal-Part-201 | mIOU: 36.9 -> 36.9 | No clear effect |
| Add factorized object and part prediction instead of concatenated category-part labels (ablation) | Multi-object multi-part parsing · Pascal-Part-58 | mIOU: 54.3 -> 61.0 | Helped |
| Add factorized object and part prediction instead of concatenated category-part labels (ablation) | Multi-object multi-part parsing · Pascal-Part-108 | mIOU: 41.3 -> 48.0 | Helped |
| Include upper/lower side attributes in the side decoder (ablation) | FLOAT · Pascal-Part | convergence | Hurt |
Closest papers by meaning
- Fine-Grained Image-Text Correspondence with Cost Aggregation for Open-Vocabulary Part Segmentation (CVPR 2025 · similarity 0.87)
- LangHOPS: Language Grounded Hierarchical Open-Vocabulary Part Segmentation (NeurIPS 2025 · similarity 0.86)
- PARTONOMY: Large Multimodal Models with Part-Level Visual Understanding (NeurIPS 2025 · similarity 0.86)
- Understanding Multi-Granularity for Open-Vocabulary Part Segmentation (NeurIPS 2024 · similarity 0.86)
- Multi-Part Object Representations via Graph Structures and Co-Part Discovery (NeurIPS 2026 · similarity 0.85)
- RESAnything: Attribute Prompting for Arbitrary Referring Segmentation (NeurIPS 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.