|
세코어로보틱스
세코어로보틱스
|
|---|
|
|
|
This demo showcases how all core 3D robotic intelligence—3D Perception, Mapping, and Localization—can run entirely from a single camera, without any LiDAR or depth sensors. 🔍 3D Perception Our model predicts full 3D structure using only images. It is trained to infer not only the visible scene but also occluded areas and the backsides of objects, enabling rich spatial understanding from pure vision. 🗺️ Mapping We generate accurate and detailed 3D maps of the environment using just camera input. No specialized 3D sensors are required. 📍 Localization Even in a “kidnapping scenario”—when the robot is suddenly moved to an unknown location—our system can re-estimate its position using only RGB images with high accuracy. ⚙️ All Running on a Single Embedded Board Every component runs efficiently on a compact embedded device, delivering real-time performance with extremely low power consumption. |
|
대표이미지 설명
세코어로보틱스 로고
|
|
|
|