Research map / Detection and segmentation / Open-set and few-shot detection
Textual Query-Driven Mask Transformer for Domain Generalized Segmentation
Byeonghyun Pak, Byeongju Woo, Sunghwan Kim, Dae-hwan Kim, Hoseong Kim
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 randomly initialized object queries with textual object queries (ablation) | tqdm (EVA02-L, GTA5 trained) · GTA5->{C,B,M} average | mIoU: 64.24 -> 64.93 | Improved |
| add text-to-pixel attention to the pixel decoder (ablation) | tqdm (EVA02-L) · GTA5->{C,B,M} average | mIoU: 64.93 -> 66.05 | Improved |
| add vision-language (pixel-text score map) regularization loss (ablation) | tqdm (EVA02-L) · GTA5->{C,B,M} average | mIoU: 63.02 -> 66.05 | Improved |
| add language regularization loss (ablation) | tqdm (EVA02-L) · GTA5->Cityscapes | mIoU: 69.34 -> 68.88 (C); 65.44 -> 66.05 (avg) | Mixed |
| add vision regularization loss (ablation) | tqdm (EVA02-L) · GTA5->{C,B,M} average | mIoU: 65.31 -> 66.05 | Improved |
| use fixed class-query matching instead of bipartite matching (ablation) | tqdm (EVA02-L) · GTA5->{C,B,M} average | mIoU: 64.19 -> 66.05 | Improved |
| use learnable prompt instead of fixed template prompt (ablation) | tqdm (EVA02-L) · GTA5->{C,B,M} average | mIoU: 65.20 -> 66.05 | Improved |
| use textual object queries instead of random queries in simple encoder-only motivating model (ablation) | CLIP ViT-B encoder with queries · unseen target domains from GTA5/SYNTHIA | mIoU | Improved |
| use tqdm instead of Rein (EVA02-CLIP) on BDD100K target | tqdm vs Rein · GTA5->BDD100K | mIoU: 60.50 -> 59.18 | Hurt |
Closest papers by meaning
- MaskViM: Domain Generalized Semantic Segmentation with State Space Models (AAAI 2025 · similarity 0.90)
- Exploring Semantic Consistency and Style Diversity for Domain Generalized Semantic Segmentation (AAAI 2025 · similarity 0.90)
- Better to Teach than to Give: Domain Generalized Semantic Segmentation via Agent Queries with Diffusion Model Guidance (ICML 2025 · similarity 0.89)
- Leveraging Depth and Language for Open-Vocabulary Domain-Generalized Semantic Segmentation (NeurIPS 2025 · similarity 0.89)
- Collaborating Foundation Models for Domain Generalized Semantic Segmentation (CVPR 2024 · similarity 0.89)
- Learning Content-Enhanced Mask Transformer for Domain Generalized Urban-Scene Segmentation (AAAI 2024 · 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.
