3.1.2 研究成果
1) 获批项目
[1] 面向煤矿井下复杂视觉环境的关键目标动态分割大模型研究,国家自然科学基金青年科学基金项目,2025.01-今,项目负责人,在研。
[2] 2019年度博士后中国科学院特别研究助理资助项目,2019.10-2021.06,项目负责人,结题。
[3] 多光谱图像云/云阴影标记数据集及智能检测算法研究与评价,中国科学院空天院横向项目,2021.11-2022.12,项目负责人,结题。
2) 代表性论文
[1] Jiao L, Huo L, Hu C, et al. Refined UNet v3: Efficient end-to-end patch-wise network for cloud and shadow segmentation with multi-channel spectral features[J]. Neural Networks, 2021, 143: 767-782.
[2] Jiao L, Huo L, Hu C, et al. Refined UNet: Unet-Based Refinement Network for Cloud and Shadow Precise Segmentation[J]. Remote Sensing, 2020, 12(12):2001.
[3] Jiao L, Huo L, Hu C, et al. Permutohedral Refined Unet: Bilateral Feature-Scalable Segmentation Network for Edge-Precise Cloud and Shadow Detection[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2024, 17: 10468-10489.
[4] Jiao L, Huo L, Hu C, et al. Refined UNet V2: End-to-End Patch-Wise Network for Noise-Free Cloud and Shadow Segmentation[J]. Remote Sensing, 2020, 12:3530.
[5] Jiao L, Huo L, Hu C, et al. Refined UNet V4: End-to-End Patch-Wise Network for Cloud and Shadow Segmentation with Bilateral Grid[J]. Remote Sensing, 2022, 14(2): 358.
[6] Jiao L, Zheng M, Tang P, et al. Towards Edge-Precise Cloud and Shadow Detection on the GaoFen-1 Dataset: A Visual, Comprehensive Investigation[J]. Remote Sensing, 2023, 15(4): 906.
[7] Jiao L, Huo L, Hu C, et al. Efficient Refined UNets: Efficient segmentation networks for boundary-aware cloud and shadow detection[J], Expert Systems with Applications, 2026, 297: 129403.
3) 论著
[1] 唐娉 等. 全球遥感数据自动化处理技术与系统架构[M]. 北京: 科学出版社, 2025. (参与撰写)
4) 开放源代码
[1] Refined UNet v3: https://github.com/jiaolobel/refined-unet-v3
[2] Refined UNet v4: https://github.com/jiaolobel/refined-unet-v4
[3] Permutohedral Refined UNet: https://github.com/jiaolobel/permutohedral-refined-unet
[4] Efficient Refined UNets: https://github.com/jiaolobel/efficient-refined-unets
[5] Efficient Global Perm Refined UNet (work in progress): https://github.com/jiaolobel/perm-refined-unet-efficient-impls
3.2.2 研究成果
1) 获批项目
[1] 面向实时运动信号分析的智能关键技术合作研究,国家级外专项目,2021.01-2022.12,项目负责人,结题。
[2] 面向生物反馈系统的智能动作识别模型及可解释性合作研究,2022年度中国-中东欧国家高校联合教育项目,2023.03-今,项目负责人,在研。
2) 代表性论文
[1] Jiao L, Gao W, Bie R, et al. Golf Guided Grad-CAM: attention visualization within golf swings via guided gradient-based class activation mapping[J]. Multimedia Tools and Applications, 2023, 83: 38481-38503.
[2] Jiao L, Bie R, Wu H, et al. Golf swing classification with multiple deep convolutional neural networks[J]. International Journal of Distributed Sensor Networks, 2018, 14(10): 1550147718802186.
[3] Jiao L, Wu H, Bie R, et al. Towards Real-Time Multi-Sensor Golf Swing Classification Using Deep CNNs[J]. Journal of Database Management (JDM), 2018, 29(3): 17-42.
3) 开放源代码
Golf Guided GradCAM: https://github.com/jiaolobel/golf-guided-gradcam
3.3.2 研究成果
1) 项目
[1] 遥感中高空间分辨率多光谱影像云遮挡神经修复技术研究,中国博士后科学基金面上项目,2020.01-2021.06,项目负责人,结题。
2) 论文
[1] Jiao L, Hu C, Huo L, et al. Guided-Pix2Pix: End-to-End Inference and Refinement Network for Image Dehazing[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2021, 14: 3052-3069.
[2] Jiao L, Hu C, Huo L, et al. Guided-Pix2Pix+: End-to-end spatial and color refinement network for image dehazing[J]. Signal Processing: Image Communication, 2022, 107: 116758.
3) 开放源代码
[1] Guided-Pix2Pix: https://github.com/jiaolobel/guided-pix2pix
[2] Guided-Pix2Pix+: https://github.com/jiaolobel/guided-pix2pixplus