Write code to add new use cases to our computer vision model - new components, wiring conditions, code violations, and safety hazards it needs to recognize and reason about.
Build and structure training data/scenarios for each new use case, and implement the scripts/pipelines needed to feed them into training.
