针对小样本深度学习技术、多模态数据源域适应技术蓬勃发展的机遇,以研发系列具有实用价值的小样本深度学习模型、域适应模型,并将其集成应用于自动驾驶领域、智慧电力、智慧遥感、火灾智能监测、网络安全监测等领域,并已取得部分处于国内前列的技术成果。
一、主持横向委托项目如下:
[1] 视觉与多模态算法模型开发;
[2] 驾驶员安全监测模型开发;
[3] 无人机视频图像电力设备缺陷检测模型开发;
[4] 视频水流速估计模型开发;
[5] 复杂环境火灾智能监测系统;
[6] 反爬虫模型开发。
二、第一发明人授权/申请专利如下:
[1] 艾加秋,黄默,王非凡,范高伟,裴志林,基于多指标卷积自编码器的PolSAR地物精细分类方法,2024-2-13,中国发明专利,ZL202111319995.4(已授权);
[2] 艾加秋,田瑞田,杨航,曹振翔,基于多层级特征深度融合的SAR图像舰船鉴别方法,2022-2-18,中国发明专利,ZL201910674382.9(已授权);
[3] 艾加秋,薛维宝,朱雅喃,基于SAR/AIS跨模态融合YOLO网络的弱小舰船检测方法,2024-6-27,中国发明专利,202410845629.X(实审);
[4] 艾加秋,朱雅喃,吴明阳,杨森,李宗豫,基于活跃态和稳态信息融合的跨域人脸表情识别方法,2024-4-2,中国发明专利,202410395073.9(实审);
[5] 艾加秋,薛维宝,朱雅喃,杨森,AIS数据辅助Transformer的双极化SAR图像海陆分割方法,2024-1-22,中国发明专利,202410088050.3(实审);
[6] 艾加秋,朱雅喃,薛维宝,基于主动学习的多目标跨域图像分类方法,2023-10-12,中国发明专利,202311320117.3(实审);
[7] 艾加秋,屈铮,一种基于中心坐标注意力模型的目标识别分类方法,2022-08-08,中国发明专利,202210943446.2(实审);
[8] 艾加秋,毛宇翔,王非凡,基于多核尺度卷积神经网络的SAR图像目标分类方法,2021-07-19,中国发明专利,202110517513.X(实审)。
三、第一作者发表代表性论文如下:
[1] Zhu Yanan, Jiaqiu Ai*, Weibao Xue, et, al. An active multi-target domain adaptation strategy: progressive class prototype rectification, IEEE Transactions on Multimedia, 2024, Accepted.
[2] Jiaqiu Ai*, Weibao Xue, Yanan Zhu, Shuo Zhuang, Congan Xu, Hao Yan*, AIS-PVT: long-time AIS data assisted pyramid vision transformer for sea-land segmentation in dual-polarization SAR imagery, IEEE Transactions on Geoscience and Remote Sensing, 2024, 62, 5220712.
[3] Jiaqiu Ai*, Yuxiang Mao, Qiwu Luo, Lu Jia, Xing Mengdao, SAR target classification using the multi-kernel-size feature fusion based convolutional neural network, IEEE Transactions on Geoscience and Remote Sensing, 2022, 60, 5214313.
[4] Jiaqiu Ai*, feifan Wang, Yuxiang Mao, Qiwu Luo, Mengdao Xing, A fine PolSAR terrain classification algorithm using the texture feature fusion based improved convolutional autoencoder, IEEE Transactions on Geoscience and Remote Sensing, 2022, 60, 5218714.
[5] Jiaqiu Ai*, Ruitian Tian, Qiwu Luo, Jing Jin, Bo Tang. Multi-scale rotation-invariant haar-like feature integrated CNN-based ship detection algorithm of multiple-target environment in SAR imagery, IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(12): 10070-10087.
[6] Jiaqiu Ai*, Zheng Qu, Zhicheng Zhao, Yong Zhang, Jun Shi, Hao Yan*, A SAR target classification algorithm based on the central coordinate attention module, IEEE Sensors Journal, 2024, 24(2): 1941-1952.
[7] Jiaqiu Ai*, Shaofan Hou, Mingyang Wu, Bin Chen, Hao Yan*, MPGSE- D-LinkNet: multiple-parameters-guided squeeze-and-excitation integrated D-LinkNet for road extraction in remote sensing imagery, IEEE Geoscience and Remote Sensing Letters, 2023, 20, 5508205.
[8] Jiaqiu Ai*, Gaowei Fan, Yuxiang Mao, Jing Jin, Mengdao Xing, He Yan, An improved SRGAN based ambiguity suppression algorithm for SAR ship target contrast enhancement, IEEE Geoscience and Remote Sensing Letters, 2022, 19, 4018705.
[9] Jiaqiu Ai*, Zhengqu, Gaowei Fan, et, al. Synthetic aperture radar ground target image generation based on improved Wasserstein generative adversarial networks with gradient penalty, Journal of Applied Remote Sensing, 2023, 17(3), 036501.