Aug 2026· 2026 6th International Conference on Advanced Algorithms and Neural Networks (AANN)· pp. 336-340· 0 citations· 17 references
Abstract
Remote sensing images are characterized by complex object and background features, which present significant challenges for ground object classification. Moreover, many common classification methods are originally designed for natural images and lack specific optimizations for remote sensing imagery. To address this, we propose a novel ground object classification method, YOLO11m-CBAM, based on YOLO11 and the Convolutional Block Attention Module (CBAM). Noting that prior works rarely discuss the optimal placement of attention modules, we first analyze the architectural characteristics of YOLO11m and the feature enhancement mechanism of CBAM. We then design several insertion schemes and conduct systematic experiments. Based on the results, we find that inserting CBAM in the middle layers of the YOLO11m backbone yields moderate improvements, while inserting it at the deepest layer achieves the best performance, which we adopt as our improved model. On the NWPU VHR-10 dataset, YOLO11m-CBAM improves mAP50-95 by 1.3% and mAP50 by 1.8% over YOLO11m. Furthermore, our model also outperforms YOLO11m on the DIOR dataset, demonstrating its generalization capability.
High-resolution remote sensing building extraction often requires large amounts of manually labeled data, and models trained on one dataset may not perform equally well on another without additional annotations. To reduce this dependency on labeled target data, we propose SIGA-UNet3+, which combines a Spectral Index Ga...
Jin-Chen Zhu, Sheng-Bo Chen, Xi-Tong Xu et al.· Italian National Conference...· 0 citations
A novel model is introduced by assessing the impacts of several YOLO object detection algorithms with the Convolutional Block Attention Module (CBAM) on aircraft detection from satellite images to demonstrate that attention mechanisms have a significant impact when used with the YOLO architecture for object detection i...
Ibrahim Aruk, Hakan Açıkgöz, Ertuğrul Doğruluk· Konya Journal of Engineering...· 0 citations
Experimental results on the ISPRS Potsdam and Vaihingen benchmark datasets show that the proposed CNN-Transformer model attains highly competitive segmentation performance while maintaining computational efficiency.
Wambugu Naftaly, Rui-Sheng Wang, Bo Guo et al.· Modeling Earth Systems and E...· 0 citations
Results verify that AARM effectively enhances the representation of small objects in complex remote sensing conditions of UAVs and achieves stable improvements in detection accuracy while maintaining high inference speed.
Yong Yang, Tian-Ci Wan, Meng-Lu Zhang· Signal, Image and Video Proc...· 0 citations
This paper suggests a practical application method that effectively detects buildings in HR aerial photos in Baghdad by creating a segmentation mask using the Mask R-CNN deep learning model. The paper used aerial 10 cm resolution images of Baghdad as the main case study, focusing on a particular urban region neighborho...
Aseel K. Hasheem, F. Abed· Iraqi Journal of Science· 0 citations
Automatic classification of military aircraft in satellite imagery is a challenging problem with a high risk of error due to the limited pixel area of targets, variations in image resolution and illumination conditions, background complexity, and strong visual similarity among classes. In this study, a deep learning ap...