Research map / Detection and segmentation / Weakly supervised detection and segmentation
Shatter and Gather: Learning Referring Image Segmentation with Text Supervision
Dongwon Kim, Namyup Kim, Cuiling Lan, Suha Kwak
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 the entity discovery module (bottom-up slot attention) (ablation) | Shatter and Gather · RefCOCO val | mIoU: 5.91 -> 34.76 | Improved |
| add the modality fusion module with C3 loss (ablation) | Shatter and Gather · RefCOCO val | mIoU: 14.65 -> 34.76 | Improved |
| add reconstruction loss (ablation) | Shatter and Gather · RefCOCO val | mIoU: 21.84 -> 34.76 | Improved |
| use slot attention in aggregation block rather than a transformer (ablation) | Shatter and Gather · RefCOCO val | mIoU: 12.61 -> 34.76 | Improved |
| add the interaction block (ablation) | Shatter and Gather · RefCOCO val | mIoU: 33.69 -> 34.76 | Improved |
| use entity slots instead of query slots (36 slots) (ablation) | Shatter and Gather · RefCOCO val | mIoU: 18.16 -> 34.76 | Improved |
| use entity slots instead of random slots (36 slots) (ablation) | Shatter and Gather · RefCOCO val | mIoU: 30.15 -> 34.76 | Improved |
| increase number of query slots from 24 to 36 (ablation) | Query slot variant · RefCOCO val | mIoU: 24.78 -> 18.16 | Hurt |
| weight slot maps by A_fuse rather than uniform averaging (ablation) | Shatter and Gather · RefCOCO val | mIoU: 8.16 -> 34.76 | Improved |
| use more entity discovery iterations (1 to 6) (ablation) | Shatter and Gather · RefCOCO val | mIoU: 28.00 -> 34.76 | Improved |
| weakly supervised model versus TSEG | Shatter and Gather · RefCOCO val | mIoU: 25.44 -> 34.76 | Improved |
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.
- GroupViT — GroupViT: Semantic Segmentation Emerges From Text Supervision (CVPR 2022) · better on 1 of 1 shared result
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
- Curriculum Point Prompting for Weakly-Supervised Referring Image Segmentation (CVPR 2024 · similarity 0.91)
- Boosting Weakly Supervised Referring Image Segmentation via Progressive Comprehension (NeurIPS 2024 · similarity 0.91)
- Prompt-Driven Referring Image Segmentation with Instance Contrasting (CVPR 2024 · similarity 0.89)
- Mask Grounding for Referring Image Segmentation (CVPR 2024 · similarity 0.89)
- CLIP as RNN: Segment Countless Visual Concepts without Training Endeavor (CVPR 2024 · similarity 0.89)
- IteRPrimE: Zero-shot Referring Image Segmentation with Iterative Grad-CAM Refinement and Primary Word Emphasis (AAAI 2025 · similarity 0.89)
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.
