研究人员开发了新的开放词汇语义分割方法，该任务涉及使用灵活的类别词汇为图像分配语义标签，而无需像素级训练数据。一种方法 LASA 聚合了 Vision Transformers 不同层的注意力图，以捕捉全局结构和局部细节，从而提高分割精度和空间连贯性。另一种方法将可微分模糊逻辑与 SAM 等基础模型集成，以改进伪标签并训练分割模型，取得了甚至超越密集监督基线的最新成果。第三种技术 Open-V 使用一种无需训练的框架，协调 SAM 和 CLIP 等模型的冻结语义先验，实现泛化少样本分割，在无需参数调整的情况下展现出强大的性能。
AI
arXiv:2606.11837v1 Announce Type: cross Abstract: Open-vocabulary scene sketch semantic segmentation aims to assign dense semantic labels to sparse line drawings based on flexible category vocabularies specified at inference time, without relying on pixel-level annotations during…
arXiv:2605.13674v2 Announce Type: replace-cross Abstract: Weakly supervised semantic segmentation (WSSS) trains dense pixel-level segmentation models from partial or coarse annotations such as bounding boxes, scribbles, or image-level tags. While recent work leverages foundation …
Open-vocabulary scene sketch semantic segmentation aims to assign dense semantic labels to sparse line drawings based on flexible category vocabularies specified at inference time, without relying on pixel-level annotations during training. Unlike natural images, sketches lack te…
arXiv cs.CV
TIER_1English(EN)·Yang Sun, Tao Wang, Anastasia Ioannou, Ge Xu·
arXiv:2606.08001v1 Announce Type: new Abstract: Semantic image segmentation assigns a predefined category label to each pixel, has achieved significant progress lately. Open-Vocabulary Segmentation (OVS) extends the segmentation task from a fixed set to an open set, enabling the …
arXiv:2606.09474v1 Announce Type: new Abstract: Generalized Few-Shot Semantic Segmentation (GFSS) has traditionally been approached as a representation-learning problem, requiring task-specific adaptation to incorporate novel classes from limited support examples. Recent foundati…
Generalized Few-Shot Semantic Segmentation (GFSS) has traditionally been approached as a representation-learning problem, requiring task-specific adaptation to incorporate novel classes from limited support examples. Recent foundation models, however, already exhibit strong open-…
