arXiv 2505.140432025-05-23cs.CV

选择性结构化状态空间用于多光谱融合小目标检测

Selective Structured State Space for Multispectral-fused Small Target Detection

Institute for AI Industry Research (AIR), Tsinghua University（人工智能产业研究院（AIR），清华大学）；National Space Science Center, Chinese Academy of Sciences（中国科学院国家空间科学中心）；School of Computer Science and Technology, University of Chinese Academy of Sciences（中国科学院大学计算机科学与技术学院）；School of Astronomy and Space Science, University of Chinese Academy of Sciences（中国科学院大学天文与空间科学学院）；School of Computing, National University of Singapore（新加坡国立大学计算机学院）

Qianqian Zhang, WeiJun Wang, Yunxing Liu, Li Zhou, Hao Zhao, Junshe An, Zihan Wang

备注：: This work was submitted to CVPR 2025, but was rejected after being reviewed by 7 reviewers. After revision, it is currently under review

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对高分辨率遥感小目标检测精度低和计算成本高的问题，提出基于Mamba的ESTD、CARG和MEPF模块，实现高效且精准的多光谱融合小目标检测。

摘要：

高分辨率遥感图像中的目标检测面临小目标识别精度低和计算成本高的挑战。Transformer架构的计算复杂度随图像分辨率呈二次方增长，而卷积神经网络（CNN）架构被迫堆叠更深的卷积层以扩大感受野，导致计算需求爆炸式增长。为解决这些计算约束，我们利用Mamba的线性复杂度来提高效率。然而，Mamba对小目标的性能下降，主要原因是小目标在图像中占据的区域有限且语义信息有限。准确识别这些小目标不仅需要Mamba的全局注意力能力，还需要精确捕捉细粒度的局部细节。为此，我们通过开发增强小目标检测（ESTD）模块和卷积注意力残差门控（CARG）模块来增强Mamba。ESTD模块增强局部注意力以捕捉细粒度细节，而CARG模块基于Mamba构建，强调空间和通道维度信息，共同提高模型捕捉小目标独特表示的能力。此外，为了突出小目标的语义表示，我们设计了一个掩码增强像素级融合（MEPF）模块用于多光谱融合，通过有效融合可见光和红外多模态信息来增强目标特征。

英文摘要：

Target detection in high-resolution remote sensing imagery faces challenges due to the low recognition accuracy of small targets and high computational costs. The computational complexity of the Transformer architecture increases quadratically with image resolution, while Convolutional Neural Networks (CNN) architectures are forced to stack deeper convolutional layers to expand their receptive fields, leading to an explosive growth in computational demands. To address these computational constraints, we leverage Mamba's linear complexity for efficiency. However, Mamba's performance declines for small targets, primarily because small targets occupy a limited area in the image and have limited semantic information. Accurate identification of these small targets necessitates not only Mamba's global attention capabilities but also the precise capture of fine local details. To this end, we enhance Mamba by developing the Enhanced Small Target Detection (ESTD) module and the Convolutional Attention Residual Gate (CARG) module. The ESTD module bolsters local attention to capture fine-grained details, while the CARG module, built upon Mamba, emphasizes spatial and channel-wise information, collectively improving the model's ability to capture distinctive representations of small targets. Additionally, to highlight the semantic representation of small targets, we design a Mask Enhanced Pixel-level Fusion (MEPF) module for multispectral fusion, which enhances target features by effectively fusing visible and infrared multimodal information.

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。