PiLoT：面向无人机自身与目标地理定位的神经像素到三维配准
PiLoT: Neural Pixel-to-3D Registration for UAV-based Ego and Target Geo-localization
把实时视频流与带地理参考的三维地图做神经像素到三维配准，用双线程引擎把地图渲染与定位解耦，并用仿真数据训练轻量网络实现零样本仿真到真实迁移，为无 GNSS 场景下的相机位姿估计提供了可复用的地图配准范式。
原文依据 PiLoT breaks this paradigm by directly registering live video stream against a geo-referenced 3D map.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论本文提出 PiLoT，一个统一框架，用于解决基于无人机的自身地理定位与目标地理定位问题。传统方法依赖解耦的流水线：用 GNSS 与视觉惯性里程计（VIO）融合来估计自身位姿，并用激光测距仪等主动传感器进行目标定位。然而，这些方法在 GNSS 拒止环境中容易失效，且带来可观的硬件成本与复杂度。PiLoT 打破这一范式，直接将实时视频流与带地理参考的三维地图进行配准。为实现鲁棒、准确且实时的性能，我们引入三项关键贡献：1）双线程引擎，将地图渲染与核心定位线程解耦，在保持无漂移精度的同时确保低延迟；2）一个带有精确几何标注（相机位姿、深度图）的大规模合成数据集，该数据集使得训练一个轻量网络成为可能，该网络能够以零样本方式从仿真泛化到真实数据；3）联合神经引导随机梯度优化器（JNGO），即使在剧烈运动下也能实现鲁棒收敛。在一系列全面的公开基准与新采集基准上的评估表明，PiLoT 优于最优方法，同时在 NVIDIA Jetson Orin 平台上以超过 25 FPS 运行。我们的代码与数据集见：https://github.com/Choyaa/PiLoT。
英文原文摘要 1376 字符
We present PiLoT, a unified framework that tackles UAV-based ego and target geo-localization. Conventional approaches rely on decoupled pipelines that fuse GNSS and Visual-Inertial Odometry (VIO) for ego-pose estimation, and active sensors like laser rangefinders for target localization. However, these methods are susceptible to failure in GNSS-denied environments and incur substantial hardware costs and complexity. PiLoT breaks this paradigm by directly registering live video stream against a geo-referenced 3D map. To achieve robust, accurate, and real-time performance, we introduce three key contributions: 1) a Dual-Thread Engine that decouples map rendering from core localization thread, ensuring both low latency while maintaining drift-free accuracy; 2) a large-scale synthetic dataset with precise geometric annotations (camera pose, depth maps). This dataset enables the training of a lightweight network that generalizes in a zero-shot manner from simulation to real data; and 3) a Joint Neural-Guided Stochastic-Gradient Optimizer (JNGO) that achieves robust convergence even under aggressive motion.Evaluations on a comprehensive set of public and newly collected benchmarks show that PiLoT outperforms state-of-the-art methods while running over 25 FPS on NVIDIA Jetson Orin platform. Our code and dataset are available at: https://github.com/Choyaa/PiLoT.