arXiv 2508.153132025-09-15cs.CVcs.AI

First RAG, Second SEG：一种用于伪装目标检测的免训练范式

First RAG, Second SEG: A Training-Free Paradigm for Camouflaged Object Detection

Nanjing University of Aeronautics and Astronautics（南京航空航天大学）

Wutao Liu, YiDan Wang, Pan Gao

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对伪装目标检测免微调且需高质量提示的难题，提出免训练的RAG-SEG范式，通过检索增强生成粗略掩码提示并利用SAM2进行精细化分割，在个人笔记本上即可取得媲美或超越SOTA的性能。

摘要：

伪装目标检测（COD）由于目标与背景高度相似，在计算机视觉中构成重大挑战。现有方法通常依赖大量训练和庞大的计算资源。虽然 Segment Anything Model (SAM) 等基础模型提供了强大的泛化能力，但它们在未经微调的情况下仍难以处理 COD 任务，且需要高质量提示才能获得良好性能。然而，手动生成此类提示成本高昂且效率低下。为应对这些挑战，我们提出了 First RAG, Second SEG (RAG-SEG)，这是一种免训练范式，将 COD 解耦为两个阶段：检索增强生成（RAG）用于生成粗略掩码作为提示，随后进行基于 SAM 的分割（SEG）以进行细化。RAG-SEG 通过无监督聚类构建紧凑的检索数据库，实现快速有效的特征检索。在推理期间，检索到的特征生成伪标签，利用 SAM2 指导精确掩码生成。我们的方法消除了传统训练的需求，同时保持了具有竞争力的性能。在基准 COD 数据集上的大量实验表明，RAG-SEG 的表现与最先进的方法持平或超越它们。值得注意的是，所有实验均在个人笔记本电脑上进行，突显了我们方法的计算效率和实用性。我们在附录中提供了进一步分析，涵盖局限性、显著目标检测扩展以及可能的改进。代码：https://github.com/Lwt-diamond/RAG-SEG。

英文摘要：

Camouflaged object detection (COD) poses a significant challenge in computer vision due to the high similarity between objects and their backgrounds. Existing approaches often rely on heavy training and large computational resources. While foundation models such as the Segment Anything Model (SAM) offer strong generalization, they still struggle to handle COD tasks without fine-tuning and require high-quality prompts to yield good performance. However, generating such prompts manually is costly and inefficient. To address these challenges, we propose \textbf{First RAG, Second SEG (RAG-SEG)}, a training-free paradigm that decouples COD into two stages: Retrieval-Augmented Generation (RAG) for generating coarse masks as prompts, followed by SAM-based segmentation (SEG) for refinement. RAG-SEG constructs a compact retrieval database via unsupervised clustering, enabling fast and effective feature retrieval. During inference, the retrieved features produce pseudo-labels that guide precise mask generation using SAM2. Our method eliminates the need for conventional training while maintaining competitive performance. Extensive experiments on benchmark COD datasets demonstrate that RAG-SEG performs on par with or surpasses state-of-the-art methods. Notably, all experiments are conducted on a \textbf{personal laptop}, highlighting the computational efficiency and practicality of our approach. We present further analysis in the Appendix, covering limitations, salient object detection extension, and possible improvements. \textcolor{blue} {Code: https://github.com/Lwt-diamond/RAG-SEG.}

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。

AI总结：

摘要：

英文摘要：

相关论文