Submission Deadline: July 13, 2026
Multi-Modal Perception and Localization/Navigation for Unmanned Systems
多模态感知与无人系统定位导航
| Chair: |
|
 |
|
| Na Lv |
|
| Shanghai Jiao Tong University, China |
|
| |
|
| Topics: |
|
- Multi-Sensor Fusion and Feature Enhancement (多模态传感器融合与特征增强)
- Tiny Target Recognition in Unstructured Environments (非结构化环境下的微小目标识别)
- High-Precision Multi-Dimensional Localization (多维空间高精度定位)
- Autonomous Localization and Mapping (SLAM) (自主定位与建图（SLAM）)
- Spatiotemporal Calibration and Kinematic Mapping (时空标定与运动学映射)
- Perception-Driven Navigation and Control (感知驱动的导航与控制)
|
| |
|
| Summary: |
|
- With the rapid progression of embodied intelligence and robotics, autonomous operations of intelligent unmanned platforms in unstructured environments face dual challenges in perception precision and spatial localization. A single sensory modality typically struggles with dynamic lighting, heavy occlusions, and varying spatial structures. This special session focuses on the deep integration of "Multi-Modal Perception" and "Localization/Navigation for Unmanned Systems." It aims to explore how heterogeneous sensors (e.g., vision, ToF, LiDAR, INS) can be leveraged to achieve environmental feature enhancement, high-precision 3D spatiotemporal mapping of tiny objects, and full-process autonomous localization and motion navigation. We welcome cutting-edge contributions regarding multi-sensor spatiotemporal calibration, cross-modal guided perception, 3D scene understanding, real-time SLAM, visual servoing, and field robot applications in smart agriculture and special inspections.
|
| |
|
- 随着具身智能与机器人技术的飞速发展，智能无人平台在非结构化环境中的自主作业面临着感知精度与空间定位的双重挑战。单一模态的传感器难以应对复杂的光照变化、密集遮挡及多变的空间结构。本分会场聚焦于“多模态感知”与“无人系统定位导航”的深度融合，旨在探讨如何利用异构传感器（如视觉、ToF、激光雷达、惯导等）实现环境特征增强、微小目标高精度三维时空映射，以及全流程的自主定位与运动导航。我们欢迎关于多传感器时空标定、跨模态引导感知、三维理解、实时SLAM、视觉伺服控制以及机器人在智能农业和特种巡检中应用的最新研究成果。
|