Teaching Spot to See: Personalized Object Learning for Assistive Robotics
Professor: Daisy Tang
Description: This project explores personalized object recognition for assistive robots, similar to enrolling a face in a facial-recognition system. A human operator will teach Boston Dynamics Spot to recognize a medication bottle from multiple viewpoints, enabling Spot to identify it later among similar objects. The scholar will develop and evaluate computer-vision and machine-learning methods under changes in viewpoint, lighting, distance, occlusion, and clutter, including rejection of uncertain matches.
Preferred Qualifications: Python
