Research
ZeroDSS: Semantic Segmentation from Class Names Alone
arXiv:2610.24405v1 Announce Type: new Abstract: Semantic segmentation depends heavily on real images with pixel-level annotations, making dense supervision a major obstacle to deployment in new domain
arXiv:2610.24405v1 Announce Type: new Abstract: Semantic segmentation depends heavily on real images with pixel-level annotations, making dense supervision a major obstacle to deployment in new domains. We study task-specific data-free semantic segmentation, where training uses generic pretrained generative priors but has access to no target-domain images, masks, or captions--only a predefined category set. We introduce Zero-Data Semantic Segmentation (ZeroDSS), a feedback-driven approach that treats synthetic supervision as an adaptively controllable training distribution rather than a direct surrogate for real data. ZeroDSS constructs a semantically structured prompt distribution through LLM-guided compositional generation and adaptive class rebalancing, then converts the prompts into image-mask pairs using diffusion attention. Its feedback-driven distribution refinement consists of a two-phase closed-loop training procedure with a single loop-back step. In Phase I, a segmenter is trained using confidence-weighted supervision and dual-scale consistency. In Phase II, model feedback is used to adaptively target underperforming categories and refine diffusion-derived masks using model predictions. After this refined initial training, a teacher-student framework is utilized in the final phase to iteratively improve the fused supervision and segmenter through exponential moving average (EMA) updates. Using synthetic training data only, ZeroDSS achieves 35.0% mIoU on MS COCO with DeepLabV3, and 66.3% and 63.5% mIoU on PASCAL VOC with DeepLabV3 and Mask2Former, respectively. These results show that explicitly constructing and iteratively refining the synthetic supervision distribution can substantially reduce the synthetic-to-real gap in dense prediction.
Related
- Do Instance Priors Help Weakly Supervised Semantic Segmentation?
- TC-ADA: One-Shot Active Domain Adaptation for Semantic Segmentation
- SADe: Sparse-Atom Support Decontamination for Few-Shot Segmentation with Weak Support Annotations
- Weakly-Supervised Lung Nodule Segmentation via Training-Free Guidance of 3D Rectified Flow
Source: arXiv cs.CV | 2026-10-20