Research map / Detection and segmentation / Open-set and few-shot detection
Semantics-Aware Dynamic Localization and Refinement for Referring Image Segmentation
Read the paperCheck 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 |
|---|---|---|---|
| Add SADLR to LAVT | LAVT · RefCOCO val | oIoU: 73.12 -> 74.24 | Helped |
| Add SADLR to LTS | LTS · RefCOCO val | oIoU: 70.84 -> 72.22 | Helped |
| Add SADLR to VLT | VLT · RefCOCO val | oIoU: 70.31 -> 71.64 | Helped |
| Add CGAN to LAVT | LAVT · RefCOCO val | oIoU: 73.12 -> 72.56 | Hurt |
| Use 2 refinement iterations instead of 3 (ablation) | SADLR with LAVT · RefCOCO val | oIoU: 74.24 -> 73.99 | Hurt |
| Use 4 refinement iterations instead of 3 (ablation) | SADLR with LAVT · RefCOCO val | oIoU: 74.24 -> 73.56 | Hurt |
| Use two convolution layers [128,512] rather than one layer [128] (ablation) | SADLR with LAVT · RefCOCO val | oIoU: 72.98 -> 74.24 | Helped |
| Retain the prior query by summing object context, rather than replacing the query (ablation) | SADLR with LAVT · RefCOCO val | oIoU: 73.19 -> 74.24 | Helped |
Closest papers by meaning
- LQMFormer: Language-aware Query Mask Transformer for Referring Image Segmentation (CVPR 2024 · similarity 0.90)
- CLIP as RNN: Segment Countless Visual Concepts without Training Endeavor (CVPR 2024 · similarity 0.88)
- Mask Grounding for Referring Image Segmentation (CVPR 2024 · similarity 0.88)
- Test-Time Adaptation of Vision-Language Models for Open-Vocabulary Semantic Segmentation (NeurIPS 2025 · similarity 0.87)
- Boosting Weakly Supervised Referring Image Segmentation via Progressive Comprehension (NeurIPS 2024 · similarity 0.87)
- SetCon: Towards Open-Ended Referring Segmentation via Set-Level Concept Prediction (NeurIPS 2026 · similarity 0.87)
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.