datasets

数 据

        经过多年积累，已形成“数据采集-数据清洗-数据标注-模型开发-实际场景应用”的研究模式，与专业的飞手团队和数据标注团队保持密切合作，可满足科研和实际应用场景数据的多样化需求, 为学术界和工业界提供智能无人系统环境感知大规模开源数据平台。
  • VisDrone：无人机目标检测和追踪基准数据集。（Detection and Tracking Meet Drones Challenge, IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021）
  • DroneVehicle：基于无人机的RGB-T车辆检测数据集。（Drone-Based RGB-Infrared Cross-Modality Vehicle Detection via Uncertainty-Aware Learning, IEEE Transactions on Circuits and Systems for Video Technology, 2022）
  • DroneCrowd：基于无人机的人群密度图估计，计数和跟踪的人群分析数据集。（Detection, Tracking, and Counting Meets Drones in Crowds: A Benchmark, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021） 
  • AnimalDrone：基于无人机的视频动物计数数据集。（Graph Regularized Flow Attention Network for Video Animal Counting From Drones, IEEE Transactions on Image Processing, 2021）
  • DroneRGBT：基于无人机的RGB-T人群计数数据集。（RGB-T Crowd Counting from Drone: A Benchmark and MMCCN Network, Proceedings of the Asian Conference on Computer Vision, 2020）
  • MultiDrone：多无人机单目标跟踪数据集。（Multi-Drone-Based Single Object Tracking With Agent Sharing Network, IEEE Transactions on Circuits and Systems for Video Technology, 2020）
  • MDMT: 多无人机多目标跟踪数据集。（Robust Multi-Drone Multi-Target Tracking to Resolve Target Occlusion: A Benchmark，2022）
code

代 码

       围绕智能无人系统环境感知，团队开发了面向深度网络结构、目标检测、目标分割、目标追踪和目标计数等领域的模型和算法。
  • ECA-Net：高效通道注意力网络。（Detection and Tracking Meet Drones Challenge, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020）
  • ASNet：智能体共享网络（多机单目标跟踪）。（Multi-Drone-Based Single Object Tracking With Agent Sharing Network, IEEE Transactions on Circuits and Systems for Video Technology, 2020）
  • STNNet：时空近邻网络（检测、追踪和计数多任务学习）。（Detection, Tracking, and Counting Meets Drones in Crowds: A Benchmark, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021）
  • UA-CMDet：不确定感知双光目标检测网络（多模态目标检测）。（Drone-Based RGB-Infrared Cross-Modality Vehicle Detection via Uncertainty-Aware Learning, IEEE Transactions on Circuits and Systems for Video Technology, 2022）
  • SLRNet：自监督低秩网络（弱半监督语义分割）。（Learning Self-Supervised Low-Rank Network for Single-Stage Weakly and Semi-Supervised Semantic Segmentation, International Journal of Computer Vision, 2022）
  • CMEDFL：动态记忆网络（单目标跟踪）。（Learning Dynamic Compact Memory Embedding for Deformable Visual Object Tracking,  IEEE Transactions on Neural Networks and Learning Systems, 2022）
  • MIA-NET: 多无人机多目标跟踪。（Robust Multi-Drone Multi-Target Tracking to Resolve Target Occlusion: A Benchmark，2022）
  • DetFusion: 检测驱动的双光融合网络（多模态目标检测）。（DetFusion: A Detection-driven Infrared and Visible Image Fusion Network，ACM MM 2022）
  • MTCP：多任务可信伪标签学习网络（半监督人群计数）。（Multi-Task Credible Pseudo-Label Learning for Semi-supervised Crowd Counting，IEEE Transactions on Neural Networks and Learning Systems, 2022）
  • MoE-Fusion：基于混合专家的多模态融合网络。（Multi-modal Gated Mixture of Local-to-Global Experts for
    Dynamic Image Fusion，ICCV 2023）
  • TC-MoA：任务自适应的混合 Adapter网络。（Task-Customized Mixture of Adapters for General Image Fusion. CVPR 2024）
      欢迎在无人机对空对地观测领域拥有数据资源但缺乏算法能力的企业和单位开展合作。