生成精确且层级嵌套的超像素，支持多尺度视觉任务
H-SPAM: Hierarchical Superpixel Anything Model
- 从细粒度分割出发，分两阶段合并区域保持对象一致性和可控融合
- 在标准数据集上精度和规则性显著优于现有层次化方法
- 支持注意力图或用户输入调控，适合需要可调结构的任务
超像素通过将像素分组为连贯区域，提供紧凑的图像表示。现有方法在分割精度上已达到瓶颈，生成的超像素形状常含噪声。此外，多数方法仅生成单一固定尺度的划分，限制了其在需要多尺度表示的视觉流程中的应用。本文提出 H-SPAM（层次化超像素任意模型），一种统一框架，用于生成准确、规则且完全嵌套的层次化超像素。从精细分区出发，基于深度特征和外部对象先验，H-SPAM 通过两阶段区域合并过程构建层级：第一阶段保留对象一致性，第二阶段允许受控的跨对象合并。层级还可通过视觉注意力图或用户输入进行调节，以更长时间保留重要区域。在标准基准上的实验表明，H-SPAM 在精度与规则性方面显著优于现有层次化方法，同时性能与大多数最新的非层次化方法相当。代码与预训练模型已开源：https://github.com/waldo-j/hspam。
原文摘要 · Abstract (English)
Superpixels offer a compact image representation by grouping pixels into coherent regions. Recent methods have reached a plateau in terms of segmentation accuracy by generating noisy superpixel shapes. Moreover, most existing approaches produce a single fixed-scale partition that limits their use in vision pipelines that would benefit multi-scale representations. In this work, we introduce H-SPAM (Hierarchical Superpixel Anything Model), a unified framework for generating accurate, regular, and perfectly nested hierarchical superpixels. Starting from a fine partition, guided by deep features and external object priors, H-SPAM constructs the hierarchy through a two-phase region merging process that first preserves object consistency and then allows controlled inter-object grouping. The hierarchy can also be modulated using visual attention maps or user input to preserve important regions longer in the hierarchy. Experiments on standard benchmarks show that H-SPAM strongly outperforms existing hierarchical methods in both accuracy and regularity, while performing on par with most recent state-of-the-art non-hierarchical methods. Code and pretrained models are available: https://github.com/waldo-j/hspam.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。