Auto Capstone: Object Detection - train YOLO model on hammer-only dataset

[ssonpatki]

Current model takes over 6 hrs to train

• Would using COE's HPC cluster make training more efficient?
• Using HPC would require completing Intro to HPC training session

Known Concerns

• error when exporting model after training is complete (likely due to internal bugs)

Goals

• export a YOLOv11 object detection test node (for at least the hammer object) that publishes detection results to a ROS2 topic
• get certified to use HPC Cluster
• explore ideas for general detection model for the unknown competition object