SAM2-UNet: SAM2用于自然和医学图像分割

最新推荐文章于 2025-03-07 15:18:29 发布

Phoenixtree_DongZhao

最新推荐文章于 2025-03-07 15:18:29 发布

阅读量2.6k

点赞数 20

分类专栏： MyDLNote-Segmentation Large Model 文章标签：计算机视觉人工智能

本文链接：https://blog.csdn.net/u014546828/article/details/141353442

版权

SAM2-UNet: Segment Anything 2 Makes Strong Encoder for Natural and Medical Image Segmentation

https://arxiv.org/pdf/2408.08870

Project page

Abstract

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 p