A public coursework archive for a Physical AI class that covered the full data-to-driving workflow: camera-based data collection, annotation, model training, and autonomous-driving algorithm design.

Overview

• Course: Data Labeling and Model Building for Physical AI
• Semester: Summer 2026
• Department: Department of Data Science, Hanyang University
• Project type: Coursework project
• Focus: camera data collection, dataset construction, annotation, object detection, model training, and autonomous driving
• Main tools: Camera data collection, annotation tools, Python, OpenCV, object-detection models

Video

https://youtu.be/CwCegRT6wLg

Slides [KOR]

자율주행_RC카_개발기.pdf

Course Workflow

The class followed the development sequence for an embodied AI system:

Camera data collection
→ dataset curation
→ annotation and quality review
→ model training and evaluation
→ autonomous-driving algorithm design

Learning and Practice

• Collected and organized camera data directly for a driving task.
• Annotated the collected dataset and trained a perception model from the labeled data.