GROOVE X 株式会社
【GROOVE X_就業型インターンシップ】 Development of Robotics Localization System
LOVOTの位置推定とマッピングシステムの最適化と、その性能を追跡するための重要な監視ツールの開発に取り組む就業型インターンシップ。多様な家庭環境で動作するLOVOTの空間認識、ナビゲーション精度、全体的な性能の継続的な向上を目指し、グラウンドトゥルースデータ収集システムの開発、移動物体や家具の識別、Visual Place Recognitionタスクの実装、Visual SLAMの最適化などに挑戦します。
This role is centered on optimizing the localization and mapping system for LOVOT and developing essential monitoring tools to track its performance. LOVOT perceives its environment using a variety of sensors, with cameras providing rich data for understanding surrounding space and people. Operating within diverse home environments presents unique challenges, making the continuous enhancement of LOVOT's spatial awareness, navigation accuracy, and overall performance critically important. In this theme, you will build and demonstrate software to improve LOVOT's localization system, or evaluate the system, on the following topics: • Develop a "ground truth" data collection system to compare LOVOT's localization with actual measurements. • Implement image recognition to identify moving objects and furniture and mark these items on the map. • Implement a Visual Place Recognition task based on neural networks. • Optimize LOVOT Visual SLAM.
- Required Skills
- • Major in Computer Science or a related technical field.
- • Programming proficiency using Python and C++.
- • Proficiency in source code management using Git.
- • Ability to communicate technical information in English, both verbally and in writing.
- Desired Skills
- • Experience with the ROS framework.
- • Strong knowledge in computer vision tasks.
- • Experience using machine learning libraries (e.g., PyTorch).
- • Basic knowledge of visual models (e.g., ViT).
- • Basic knowledge of SLAM (Simultaneous Localization and Mapping).
- • Experience with code review processes and development within a Continuous Integration (CI) environment.