引入上下文信息和Attention Gate的GUS-YOLO遥感目标检测算法

doi:10.3778/j.issn.1673-9418.2305005

摘要/Abstract

摘要： 目前基于通用YOLO系列的遥感目标检测算法存在并未充分利用图像的全局上下文信息，在特征融合金字塔部分并未充分考虑缩小融合特征之间的语义鸿沟、抑制冗余信息干扰的缺点。在结合YOLO算法优点的基础上提出GUS-YOLO算法，其拥有一个能够充分利用全局上下文信息的骨干网络Global Backbone。除此之外，该算法在融合特征金字塔自顶向下的结构中引入Attention Gate模块，可以突出必要的特征信息，抑制冗余信息。另外，为Attention Gate模块设计了最佳的网络结构，提出了网络的特征融合结构U-Net。最后，为克服ReLU函数可能导致模型梯度不再更新的问题，该算法将Attention Gate模块的激活函数升级为可学习的SMU激活函数，提高模型鲁棒性。在NWPU VHR-10遥感数据集上，该算法相较于YOLOV7算法取得宽松指标mAP0.50 1.64个百分点和严格指标mAP0.75 9.39个百分点的性能提升。相较于目前主流的七种检测算法，该算法取得较好的检测性能。

关键词: 遥感图像, Global Backbone, Attention Gate, SMU, U-neck

Abstract: At present, there are still some problems in the remote sensing target detection algorithm based on the general YOLO (you only look once) series, such as not making full use of the global context information of the image, not narrowing the semantic gap in the feature fusion pyramid part, and not suppressing the interference of redundant information. On the basis of combining the advantages of YOLO algorithms, this paper proposes GUS-YOLO (network of global context extraction unit and attention gate-based YOLOS) algorithm. It has a backbone network Global Backbone that can make full use of global context information. Other than that, this algorithm introduces the Attention Gate module into the top-down structure of the fused feature pyramid, which can emphasize the necessary feature information and suppress redundant information. Furthermore, this paper designs the best network structure for the Attention Gate module and proposes the feature fusion structure U-Net of proposed network. Finally, because the ReLU activation function may lead to the problem that the model gradient is no longer updated, the Attention Gate module uses a learnable SMU (smooth maximum unit) activation function, which can improve the robustness of the model. On the NWPU VHR-10 remote sensing dataset, this algorithm achieves 1.64 percentage points and 9.39 percentage points performance improvement on mAP0.50 and mAP0.75 respectively compared with YOLOV7. Compared with the current 7 mainstream detection algorithms, this algorithm achieves better detection performance.

Key words: remote sensing image, Global Backbone, Attention Gate, SMU, U-neck

张华卫, 张文飞, 蒋占军, 廉敬, 吴佰靖. 引入上下文信息和Attention Gate的GUS-YOLO遥感目标检测算法[J]. 计算机科学与探索, 2024, 18(2): 453-464.

ZHANG Huawei, ZHANG Wenfei, JIANG Zhanjun, LIAN Jing, WU Baijing. GUS-YOLO Remote Sensing Target Detection Algorithm Introducing Context Information and Attention Gate[J]. Journal of Frontiers of Computer Science and Technology, 2024, 18(2): 453-464.

参考文献

[1] 汪西莉, 梁敏, 刘涛. 特征增强的单阶段遥感图像目标检测模型[J]. 西安电子科技大学学报, 2022, 49(3): 160-170.
WANG X L, LIANG M, LIU T. Feature enhanced single-stage remote sensing image object detection model[J]. Journal of Xidian University, 2022, 49(3): 160-170.
[2] 李坤亚, 欧鸥, 刘广滨, 等. 改进YOLOv5的遥感图像目标检测算法[J]. 计算机工程与应用, 2023, 59(9): 207-214.
LI K Y, OU O, LIU G B, et al. Target detection algorithm of remote sensing image based on improved YOLOv5[J]. Journal of Computer Engineering and Applications, 2023, 59(9): 207-214.
[3] 李阿标, 郭浩, 戚畅, 等. 复杂背景下遥感图像密集目标检测[J]. 计算机工程与应用, 2023, 59(8): 247-253.
LI A B, GUO H, QI C, et al. Dense object detection in remote sensing images under complex background[J]. Computer Engineering and Applications, 2023, 59(8): 247-253.
[4] CARION N, MASSA F, SYNNAEVE G, et al. End-to-end object detection with transformers[C]//Proceedings of the 16th European Conference on Computer Vision. Cham: Springer, 2020: 213-229.
[5] ZHU X, SU W, LU L, et al. Deformable DETR: deformable transformers for end-to-end object detection[J]. arXiv:2010. 04159, 2020.
[6] LIU S, LI F, ZHANG H, et al. Dab-DETR: dynamic anchor boxes are better queries for DETR[J]. arXiv:2201.12329, 2022.
[7] GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition, Columbus, Jun 23-28, 2014. Washington: IEEE Computer Society, 2014: 580-587.
[8] GIRSHICK R. Fast R-CNN[C]//Proceedings of the 2015 IEEE International Conference on Computer Vision, Santiage, Dec 7-13, 2015. Washington: IEEE Computer Society,2015: 1440-1448.
[9] ZHOU X, GIRDHAR R, JOULIN A, et al. Detecting twenty-thousand classes using image-level supervision[C]//Proceedings of the 17th European Conference on Computer Vision, Tel Aviv, Oct 23-27, 2022. Cham: Springer, 2022: 350-368.
[10] CHEN S, SUN P, SONG Y, et al. DiffusionDet: diffusion model for object detection[J]. arXiv:2211.09788, 2022.
[11] TAN M, PANG R, LE Q V. EfficientDet: scalable and efficient object detection[C]//Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Jun 13-19, 2020. Piscataway: IEEE, 2020: 10781-10790.
[12] GE Z, LIU S, WANG F, et al. YOLOX: exceeding YOLO series in 2021[J]. arXiv:2107.08430, 2021.
[13] LI C, LI L, JIANG H, et al. YOLOv6: a single-stage object detection framework for industrial applications[J]. arXiv:2209.02976, 2022.
[14] WANG C Y, BOCHKOVSKIY A, LIAO H. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for realtime object detectors[J]. arXiv:2207.02696, 2022.
[15] 朱煜, 方观寿, 郑兵兵, 等. 基于旋转框精细定位的遥感目标检测方法研究[J]. 自动化学报, 2023, 49(2): 10-17.
ZHU Y, FANG G S, ZHENG B B, et al. Research on detection method of refined rotated boxes in remote sensing[J]. Acta Automatica Sinica, 2023, 49(2): 10-17.
[16] NAIR V, HINTON G E. Rectified linear units improve restricted Boltzmann machines[C]//Proceedings of the 27th International Conference on Machine Learning, Haifa, Jun 21-24, 2010: 807-814.
[17] MAAS A L, HANNUN A Y, NG A Y. Rectifier nonlinearities improve neural network acoustic models[C]//Proceedings of the 30th International Conference on Machine Learning, Atlanta, Jun 16-21, 2013.
[18] CLEVERT D A, UNTERTHINER T, HOCHREITER S. Fast and accurate deep network learning by exponential linear units (ELUs)[J]. arXiv:1511.07289, 2015.
[19] RAMACHANDRAN P, ZOPH B, LE Q V. Searching for activation functions[J]. arXiv:1710.05941, 2017.
[20] HENDRYCKS D, GIMPEL K. Gaussian error linear units (GELUs)[J]. arXiv:1606.08415, 2016.
[21] BISWAS K, KUMAR S, BANERJEE S, et al. SMU: smooth activation function for deep networks using smoothing maximum technique[J]. arXiv:2111.04682, 2021.
[22] OKTAY O, SCHLEMPER J, FOLGOC L L, et al. Attention U-Net: learning where to look for the pancreas[J]. arXiv:1804.03999, 2018.
[23] ZHU Y, ZHAO C, WANG J, et al. CoupleNet: coupling global structure with local parts for object detection[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision, Venice, Oct 22-29, 2017. Washington: IEEE Computer Society, 2017: 4126-4134.
[24] DAI J, LI Y, HE K, et al. R-FCN: object detection via region-based fully convolutional networks[C]//Advances in Neural Information Processing Systems 29, Barcelona, Dec 5-10, 2016: 379-387.
[25] TAN M, LE Q. EfficientNet: rethinking model scaling for convolutional neural networks[C]//Proceedings of the 2019 International Conference on Machine Learning, Long Beach, Jun 9-15, 2019: 6105-6114.
[26] TAN M, LE Q. EfficientNetv2: smaller models and faster training[C]//Proceedings of the 38th International Conference on Machine Learning, Jul 18-24, 2021: 10096-10106.
[27] LIU Z, LIN Y, CAO Y, et al. Swin transformer: hierarchical vision transformer using shifted windows[C]//Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision, Montreal, Oct 10-17, 2021. Piscataway: IEEE, 2021: 10012-10022.
[28] LIU Z, MAO H, WU C Y, et al. A convnet for the 2020s[C]//Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, Jun 18-24, 2022. Piscataway: IEEE, 2022: 11966-11976.
[29] ZHAO G, GE W, YU Y. GraphFPN: graph feature pyramid network for object detection[C]//Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision, Jun 19-25, 2021. Piscataway: IEEE, 2021: 2763-2772.
[30] DING X, ZHANG X, MA N, et al. RepVGG: making VGG-style convnets great again[C]//Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Jun 19-25, 2021. Piscataway: IEEE, 2021: 13733-13742.
[31] JOCHER G, CHAURASIA A, STOKEN A, et al. Ultralytics/YOLOV5: v6.1-TensorRT, TensorFlow edge TPU and OpenVINO export and inference[Z]. Zenodo, 2022.
[32] CAO Y, XU J, LIN S, et al. GCNet: non-local networks meet squeeze-excitation networks and beyond[C]//Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision, Seoul, Oct 27-Nov 3, 2019. Piscataway: IEEE, 2019: 1971-1980.
[33] WANG X, GIRSHICK R, GUPTA A, et al. Non-local neural networks[C]//Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, Jun 18-23, 2018. Washington: IEEE Computer Society, 2018: 7794-7803.
[34] HU J, SHEN L, SUN G. Squeeze-and-excitation networks[C]//Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, Jun 18-23, 2018. Washington: IEEE Computer Society, 2018: 7132-7141.
[35] DEVLIN J, CHANG M W, LEE K, et al. BERT: pretraining of deep bidirectional transformers for language understanding[J]. arXiv:1810.04805, 2018.
[36] BROWN T, MANN B, RYDER N, et al. Language models are few-shot learners[C]//Advances in Neural Information Processing Systems 33, Dec 6-12, 2020: 1877-1901.
[37] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Advances in Neural Information Processing Systems 30, Long Beach, Dec 4-9, 2017: 5998-6008.
[38] WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C]//Proceedings of the 15th European Conference on Computer Vision, Munich, Sep 8-14, 2018. Cham: Springer, 2018: 3-19.
[39] LIU Y, SHAO Z, TENG Y, et al. NAM: normalization-based attention module[J]. arXiv:2111.12419, 2021.
[40] LIU H, LIU F, FAN X, et al. Polarized self-attention: towards high-quality pixel-wise regression[J]. arXiv:2107.00782, 2021.