SAM2-UNet：Segment Anything 2 为自然和医学图像分割构建强大编码器
SAM2-UNet: Segment Anything 2 Makes Strong Encoder for Natural and Medical Image Segmentation
AI总结：
SAM2-UNet通过结合SAM2的Hiera主干和U型解码器，为自然和医学图像分割提供强大编码器，实现了高效且无需复杂调整的多任务分割性能。
摘要：
图像分割在视觉理解中扮演着重要角色。最近，新兴的视觉基础模型在各种任务上持续取得优异表现。在这一成功的基础上，本文证明了Segment Anything Model 2（SAM2）可以作为U型分割模型的强大编码器。我们提出了一种简单但有效的框架，称为SAM2-UNet，用于多用途图像分割。具体来说，SAM2-UNet采用SAM2的Hiera主干作为编码器，而解码器使用经典U型设计。此外，将适配器插入到编码器中，以允许参数高效的微调。在各种下游任务上的初步实验，如伪装物体检测、显著物体检测、海洋动物分割、镜像检测和息肉分割，表明我们的SAM2-UNet可以简单地超越现有的专门状态-of-the-art方法，无需额外的花哨功能。项目页面：\url{https://github.com/WZH0120/SAM2-UNet}。
英文摘要：
Image segmentation plays an important role in vision understanding. Recently, the emerging vision foundation models continuously achieved superior performance on various tasks. Following such success, in this paper, we prove that the Segment Anything Model 2 (SAM2) can be a strong encoder for U-shaped segmentation models. We propose a simple but effective framework, termed SAM2-UNet, for versatile image segmentation. Specifically, SAM2-UNet adopts the Hiera backbone of SAM2 as the encoder, while the decoder uses the classic U-shaped design. Additionally, adapters are inserted into the encoder to allow parameter-efficient fine-tuning. Preliminary experiments on various downstream tasks, such as camouflaged object detection, salient object detection, marine animal segmentation, mirror detection, and polyp segmentation, demonstrate that our SAM2-UNet can simply beat existing specialized state-of-the-art methods without bells and whistles. Project page: \url{https://github.com/WZH0120/SAM2-UNet}.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
