Segment anything model (SAM) 在常见视觉场景中展现出了卓越的泛化能力，但在理解专门数据方面能力不足。近期，一些方法将参数高效技术与任务特定设计相结合，以在特定任务上微调 SAM。然而，这些方法严重依赖手工制作、复杂且任务特定的设计以及前/后处理，才能在下游任务中达到可接受的性能。因此，这严重限制了其在其他下游任务上的泛化能力。为了解决这一问题，我们提出了一个简单且统一的框架，即 SU-SAM，该框架能够利用参数高效技术轻松高效地微调 SAM 模型，同时保持对各种下游任务的出色泛化能力。SU-SAM 不需要任何任务特定的设计，旨在显著提升类 SAM 模型对表现不佳场景的适应性。具体而言，我们将不同方法的参数高效模块抽象为框架中的基本设计元素。此外，我们提出了 SU-SAM 的四种变体，即串联、并联、混合和 LoRA 结构。我们在九个数据集和六个下游任务上进行了全面实验以验证 SU-SAM 的有效性，包括医学图像分割、伪装目标检测、显著目标分割、表面缺陷分割、复杂目标形状和阴影掩码。我们的实验结果表明，与最先进的方法相比，SU-SAM 取得了具有竞争力或更优的准确率。此外，我们提供了深入分析，强调了 SU-SAM 内不同参数高效设计的有效性。另外，我们提出了一个通用模型和基准，展示了 SU-SAM 同时跨越所有多样化数据集的泛化能力。
英文摘要：
Segment anything model (SAM) has demonstrated excellent generalizability in common vision scenarios, yet falling short of the ability to understand specialized data. Recently, several methods have combined parameter-efficient techniques with task-specific designs to fine-tune SAM on particular tasks. However, these methods heavily rely on handcraft, complicated, and task-specific designs, and pre/post-processing to achieve acceptable performances on downstream tasks. As a result, this severely restricts generalizability to other downstream tasks. To address this issue, we present a simple and unified framework, namely SU-SAM, that can easily and efficiently fine-tune the SAM model with parameter-efficient techniques while maintaining excellent generalizability toward various downstream tasks. SU-SAM does not require any task-specific designs and aims to improve the adaptability of SAM-like models significantly toward underperformed scenes. Concretely, we abstract parameter-efficient modules of different methods into basic design elements in our framework. Besides, we propose four variants of SU-SAM, i.e., series, parallel, mixed, and LoRA structures. Comprehensive experiments on nine datasets and six downstream tasks to verify the effectiveness of SU-SAM, including medical image segmentation, camouflage object detection, salient object segmentation, surface defect segmentation, complex object shapes, and shadow masking. Our experimental results demonstrate that SU-SAM achieves competitive or superior accuracy compared to state-of-the-art methods. Furthermore, we provide in-depth analyses highlighting the effectiveness of different parameter-efficient designs within SU-SAM. In addition, we propose a generalized model and benchmark, showcasing SU-SAM's generalizability across all diverse datasets simultaneously.
SAM Fails to Segment Anything? -- SAM-Adapter: Adapting SAM in Underperformed Scenes: Camouflage, Shadow, Medical Image Segmentation, and More
KOKONI, Moxin (Huzhou) Tech. Co., LTD（莫西（湖州）科技有限公司KOKONI）；Beihang University（北京航空航天大学）；Zhejiang University（浙江大学）；Huzhou University（湖州师范学院）；Singapore University of Technology and Design（新加坡科技设计大学）
Multi-scale Contrastive Adaptor Learning for Segmenting Anything in Underperformed Scenes
School of Automation and Electrical Engineering, University of Science and Technology Beijing（北京科技大学自动化与电气工程学院）；Alibaba DAMO Academy（阿里巴巴达摩院）；College of Computer Science and Technology, Zhejiang University（浙江大学计算机科学与技术学院）
SAM3-UNet: Simplified Adaptation of Segment Anything Model 3
Sun Yat-sen University（中山大学）；Jiangxi Normal University（江西师范大学）；Hainan University（海南大学）；Dalian University of Technology（大连理工大学）；Chizhou University（池州学院）；Wuhan University（武汉大学）
Nanyang Technological University（南洋理工大学）；SenseTime Research（商汤科技研究院）；Carnegie Mellon University（卡内基梅隆大学）；Mohamed bin Zayed University of Artificial Intelligence（穆罕默德·本·扎耶德人工智能大学）
AI-SAM: Automatic and Interactive Segment Anything Model
The Pennsylvania State University（宾夕法尼亚州立大学）；Huck Institutes of the Life Sciences（哈克生命科学研究所）；Northwestern Memorial Hospital（西北纪念医院）；Northwestern University（西北大学）
Biomedical SAM 2: Segment Anything in Biomedical Images and Videos
University of Georgia（佐治亚大学）；Massachusetts General Hospital（麻省总医院）；Harvard Medical School（哈佛医学院）；Lehigh University（理海大学）；Northeastern University（东北大学）；University of Notre Dame（圣母大学）
On Efficient Variants of Segment Anything Model: A Survey
School of Computer Science and Engineering（计算机科学与工程学院）；School of Computing and Communications（计算与通信学院）；School of Computer Science and Technology（计算机科学与技术学院）
SAMRS: Scaling-up Remote Sensing Segmentation Dataset with Segment Anything Model
Wuhan University（武汉大学）；The University of Sydney（悉尼大学）；National Engineering Research Center of Speech and Language Information Processing（国家语音及语言信息处理工程技术研究中心）
Segment Anything Model Can Not Segment Anything: Assessing AI Foundation Model's Generalizability in Permafrost Mapping
Arizona State University（亚利桑那州立大学）；Woodwell Climate Research Center（伍德韦尔气候研究中心）；University of Connecticut（康涅狄格大学）；U.S. Geological Survey（美国地质调查局）；University of California, Santa Barbara（加利福尼亚大学圣巴巴拉分校）；University of Illinois at Urbana-Champaign（伊利诺伊大学香槟分校）
本文评估了以 Meta SAM 为代表的视觉基础模型在永久冻土地貌分割中的表现，通过多种提示策略和微调测试其零样本能力与领域适应性，并指出其在挑战性地理空间任务中仍需改进。
Generalization vs. Specialization: Evaluating Segment Anything Model (SAM3) Zero-Shot Segmentation Against Fine-Tuned YOLO Detectors
Cornell University, Department of Biological and Environmental Engineering（康奈尔大学生物与环境工程系）；University of the Peloponnese, Department of Informatics and Telecommunications（希腊皮埃里亚大学信息与电信系）
SAM2-UNet: Segment Anything 2 Makes Strong Encoder for Natural and Medical Image Segmentation
School of Computer Science and Engineering, Sun Yat-sen University（中山大学计算机科学与工程学院）；School of Computer and Information Engineering, Jiangxi Normal University（江西师范大学计算机与信息工程学院）；The Chinese University of Hong Kong (Shenzhen)（香港中文大学（深圳））；Tianjin University（天津大学）；Monash University（墨尔本大学）；Pazhou Lab（琶洲实验室）；Smart Hospital Research Institute, Peking University Shenzhen Hospital（北京大学深圳医院智能医院研究所）
SAM2-Adapter: Evaluating & Adapting Segment Anything 2 in Downstream Tasks: Camouflage, Shadow, Medical Image Segmentation, and More
KOKONI, Moxin (Huzhou) Tech. Co., LTD（科科尼（湖州）魔芯科技有限公司）；School of Information Engineering, Huzhou University（湖州大学信息工程学院）；School of Information Science and Technology, University of Science and Technology of China（中国科学技术大学信息科学技术学院）；School of Software Technology, Zhejiang University（浙江大学软件学院）；College of Computer Science and Technology, Zhejiang University（浙江大学计算机科学与技术学院）；Information Systems Technology and Design Pillar, Singapore University of Technology and Design（新加坡科技设计大学信息系统技术与设计支柱学院）
TASAM: Terrain-and-Aware Segment Anything Model for Temporal-Scale Remote Sensing Segmentation
University of Alabama at Birmingham（阿拉巴马大学伯明翰分校）；Wuhan University（武汉大学）；Dalian University of Technology（大连理工大学）；University of Chinese Academy of Sciences（中国科学院大学）
CAT-SAM: Conditional Tuning for Few-Shot Adaptation of Segment Anything Model
The University of Tokyo（东京大学）；Nara Institute of Science and Technology（奈良科学技术研究所）；Nanyang Technological University（南洋理工大学）；Waseda University（早稻田大学）；Zhejiang University of Technology（浙江工业大学）；University of Chinese Academy of Sciences（中国科学院大学）；RIKEN AIP（理化学研究所人工智能项目）
Compositional Oil Spill Detection Based on Object Detector and Adapted Segment Anything Model from SAR Images
The Hong Kong Polytechnic University（香港理工大学）；Research Institute for Sustainable Urban Development, The Hong Kong Polytechnic University（香港理工大学可持续城市发展研究院）
Enhancing the Reliability of Segment Anything Model for Auto-Prompting Medical Image Segmentation with Uncertainty Rectification
Artificial Intelligence Innovation and Incubation Institute, Fudan University（复旦大学人工智能创新与孵化研究院）；Shanghai Academy of Artificial Intelligence for Science（上海科学人工智能研究院）
MergeSAM: Unsupervised change detection of remote sensing images based on the Segment Anything Model
School of Geography and Planning, Sun Yat-sen University（中山大学地理科学与规划学院）；Unit 33269 of the PLA（中国人民解放军33269部队）；Intelligent Science & Technology Academy Limited of CASIC（中国航天科工集团智能科技研究院有限公司）
Medical SAM Adapter: Adapting Segment Anything Model for Medical Image Segmentation
University of Oxford（牛津大学）；National University of Singapore（新加坡国立大学）；University of Alberta（阿尔伯塔大学）；A*STAR（新加坡科技研究局）；Carnegie Mellon University（卡内基梅隆大学）；Singapore Eye Research Institute（新加坡眼科研究所）；Mohamed bin Zayed University of Artificial Intelligence（穆罕默德·本·扎耶德人工智能大学）；University of Hongkong（香港大学）
本文提出Medical SAM Adapter（Med-SA），通过轻量适配技术将医学知识融入SAM，利用SD-Trans和HyP-Adpt适配2D模型至3D医学图像，在17个任务上以仅2%参数更新超越SOTA方法。
SAM3-Adapter: Efficient Adaptation of Segment Anything 3 for Camouflage Object Segmentation, Shadow Detection, and Medical Image Segmentation
KOKONI, Moxin (Huzhou) Tech. Co., LTD（摩西（湖州）科技有限公司）；College of Computer Science and Technology, Zhejiang University（浙江大学计算机科学与技术学院）；School of Computer Science and Engineering, Nanjing University of Science and Technology（南京理工大学计算机科学与工程学院）；School of Information Engineering, Huzhou University（湖州大学信息工程学院）；School of Electrical and Electronic Engineering, Nanyang Technological University（新加坡南洋理工大学电子与电气工程学院）；College of Computing and Data Science, Nanyang Technological University（新加坡南洋理工大学计算与数据科学学院）；School of Information Science and Technology, University of Science and Technology of China（中国科学技术大学信息科学与技术学院）
SAM-6D: Segment Anything Model Meets Zero-Shot 6D Object Pose Estimation
DexForce Co. Ltd.（德克斯福斯有限公司）；School of Data Science, The Chinese University of Hong Kong, Shenzhen（香港中文大学（深圳）数据科学学院）；South China University of Technology（华南理工大学）
Adapting Segment Anything Model for Change Detection in HR Remote Sensing Images
Information Engineering University（信息工程大学）；Nanjing University of Information Science and Technology（南京信息工程大学）；ETH Zurich（苏黎世联邦理工学院）；University of Trento（特伦托大学）
