Research map / Detection and segmentation / Open-set and few-shot detection
PCA-Seg: Revisiting Cost Aggregation for Open-Vocabulary Semantic and Part Segmentation
Jianjian Yin, Tao Chen, Yi Chen, Gensheng Pei, Xiangbo Shu, Yazhou Yao, Fumin Shen
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 |
|---|---|---|---|
| replace serial spatial-then-class aggregation with parallel aggregation (ablation) | PCA-Seg (ViT-B/16) · A-150 | mIoU: 36.3 -> 38.1 | Improved |
| replace serial with parallel aggregation (ablation) | PCA-Seg (ViT-B/16) · PAS-20b | mIoU: 81.3 -> 82.7 | Improved |
| add cost aggregation (serial) to the cost volume (ablation) | CAT-Seg-style model · A-150 | mIoU: 34.4 -> 36.3 | Improved |
| parallelize aggregation with a single conv (no EPL/FOD) (ablation) | DeCLIP-style serial model · PAS-20b | mIoU: 81.3 -> 81.1 | No clear effect |
| add EPL module to the parallel baseline (ablation) | PCA-Seg · A-150 | mIoU: 36.4 -> 37.2 | Improved |
| add FOD strategy to the parallel baseline (ablation) | PCA-Seg · A-150 | mIoU: 36.4 -> 36.9 | Improved |
| add both EPL and FOD to the parallel baseline (ablation) | PCA-Seg · A-150 | mIoU: 36.4 -> 38.1 | Improved |
| add EPL and FOD to a serial architecture (ablation) | DeCLIP-style serial model · A-150 | mIoU: 36.3 -> 36.7 | Improved |
| use coefficient mapper instead of convolution for expert integration (ablation) | PCA-Seg · PC-59 | mIoU: 60.7 -> 61.2 | Improved |
| use coefficient mapper instead of average/addition fusion (ablation) | PCA-Seg · PC-59 | mIoU: 60.9 -> 61.2 | Improved |
| increase expert blocks from 1 to 4 (ablation) | PCA-Seg · A-150 | mIoU: 36.8 -> 37.2 | Improved |
| increase expert blocks from 4 to 5 (ablation) | PCA-Seg · A-150 | mIoU: 37.2 -> 36.9 | Hurt |
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.
- DeCLIP — DeCLIP: Decoupled Learning for Open-Vocabulary Dense Perception (CVPR 2025) · better on 9 of 9 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
- CAT-Seg: Cost Aggregation for Open-Vocabulary Semantic Segmentation (CVPR 2024 · similarity 0.91)
- Emergent Open-Vocabulary Semantic Segmentation from Off-the-shelf Vision-Language Models (CVPR 2024 · similarity 0.89)
- Exploring Simple Open-Vocabulary Semantic Segmentation (CVPR 2025 · similarity 0.89)
- Fine-Grained Image-Text Correspondence with Cost Aggregation for Open-Vocabulary Part Segmentation (CVPR 2025 · similarity 0.88)
- Test-Time Adaptation of Vision-Language Models for Open-Vocabulary Semantic Segmentation (NeurIPS 2025 · similarity 0.88)
- DPSeg: Dual-Prompt Cost Volume Learning for Open-Vocabulary Semantic Segmentation (CVPR 2025 · similarity 0.88)
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.
