Research map / Detection and segmentation / Open-set and few-shot detection
ReME: A Data-Centric Framework for Training-Free Open-Vocabulary Segmentation
Xiwei Xuan, Ziquan Deng, Kwan-Liu Ma
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-based filtering to the base reference set (ablation) | ReME · VOC-20 | mIoU: 70.03 -> 91.10 | Improved |
| add semantic enriching to the base reference set (ablation) | ReME · VOC-20 | mIoU: 70.03 -> 79.50 | Improved |
| combine group-based filtering and semantic enriching (ablation) | ReME · A-150 | mIoU: 22.03 -> 26.13 | Improved |
| filter by intra-modal visual-feature similarity instead of global CLIP score (ablation) | ReME data cleaning · VOC-20 | mIoU: 79.34 -> 92.34 | Improved |
| use group-based CLIP-score filtering instead of global CLIP-score filtering (ablation) | ReME data cleaning · VOC-20 | mIoU: 79.34 -> 80.05 | No clear effect |
| use LLaVA descriptions instead of BLIP-2 or GT captions (ablation) | ReME · COCO Object | mIoU: 37.64 -> 45.50 | Improved |
| use DINOv2 instead of CLIP as visual encoder (ablation) | ReME · PC-59 | mIoU: 36.51 -> 43.65 | Improved |
| upgrade encoder from ViT-B to ViT-L (ablation) | ReME · PC-59 | mIoU: 43.65 -> 44.89 | Improved |
| replace superpixel segmenter with SAM (ablation) | ReME · PC-59 | mIoU: 44.89 -> 53.10 | Improved |
| raise drop ratio or k_sim beyond default 30 (ablation) | ReME · COCO Stuff 1k validation images | mIoU | Hurt |
| build reference set from images with enhancement instead of synthetic images from captions | ReME vs FreeDA · PC-459 | seconds/image: 4.89 -> 0.31 | Improved |
Later papers that compared with this method
1 later accepted paper reports ReME as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- CorrCLIPv2: Harnessing Unlabeled Images for Open-Vocabulary Semantic Segmentation (IJCV 2026 · better on 1 of 1 shared result)
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.
- CorrCLIP — CorrCLIP: Reconstructing Patch Correlations in CLIP for Open-Vocabulary Semantic Segmentation (ICCV 2025) · better on 2 of 2 shared results
- ProxyCLIP — ProxyCLIP: Proxy Attention Improves CLIP for Open-Vocabulary Segmentation (ECCV 2024) · 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
- Emergent Open-Vocabulary Semantic Segmentation from Off-the-shelf Vision-Language Models (CVPR 2024 · similarity 0.92)
- Transferable and Principled Efficiency for Open-Vocabulary Segmentation (CVPR 2024 · similarity 0.92)
- Training-free Open-Vocabulary Semantic Segmentation via Diverse Prototype Construction and Sub-region Matching (AAAI 2025 · similarity 0.91)
- CLIP as RNN: Segment Countless Visual Concepts without Training Endeavor (CVPR 2024 · similarity 0.91)
- Open-Vocabulary Segmentation with Semantic-Assisted Calibration (CVPR 2024 · similarity 0.91)
- Training-Free Open-Vocabulary Segmentation with Offline Diffusion-Augmented Prototype Generation (CVPR 2024 · similarity 0.90)
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.
