Innovation Forum on 3D Vision Perception Technology in Autonomous Driving and Embodied Intelligence
自动驾驶与具身智能中的3D视觉感知技术创新论坛
| Chair: |
Co-chair: |
 |
 |
| Chunyu Lin |
Changsheng Li |
| Beijing Jiaotong University, China |
Beijing Institute of Technology, China |
| |
|
| Summary: |
|
- The forum centers on the core technical bottlenecks and industrial pain points of 3D visual perception, specifically covering key research directions including multi-sensor fusion perception, large model-driven 3D semantic understanding, joint modeling of point clouds and images, 3D reconstruction based on Neural Radiance Field (NeRF) and Gaussian Splatting (GS), BEV/Occupancy networks, end-to-end integrated perception and decision-making, environmental comprehension and interactive perception for embodied robots, as well as low-power edge deployment of 3D perception.
|
| |
|
- 论坛围绕 3D 视觉感知的核心技术瓶颈与产业痛点，具体包括：多传感器融合感知、大模型驱动的 3D 语义理解、点云与图像联合建模、神经辐射场（NeRF）/ 高斯溅射（GS）三维重建、BEV/Occupancy 占用网络、端到端感知决策一体化、具身机器人的环境理解与交互感知、低功耗边缘端 3D 感知部署等关键方向。论坛围绕 3D 视觉感知的核心技术瓶颈与产业痛点，具体包括：多传感器融合感知、大模型驱动的 3D 语义理解、点云与图像联合建模、神经辐射场（NeRF）/ 高斯溅射（GS）三维重建、BEV/Occupancy 占用网络、端到端感知决策一体化、具身机器人的环境理解与交互感知、低功耗边缘端 3D 感知部署等关键方向。
|