A marina has forty hull photos tagged barnacle or clean and wants a custom SageMaker image trainer this week. What should the ML engineer decide first?
Select an answer to reveal the explanation.
Short Explanation
Think of forty hull photos tagged barnacle or clean, and a custom SageMaker image trainer this week. That labeled set is too thin. Use a managed service such as Amazon Rekognition. Forty shots do not make a robust custom model by Friday.
Full Explanation
Assess labeled volume before committing to custom training. Forty photos are too thin for a from-scratch SageMaker image trainer, so a managed vision service such as Amazon Rekognition is the feasible path. Launching Image Classification on that folder, skipping the check, and Polly miss the volume constraint.