Face is the correct option to build a facial recognition system that identifies named staff members in photos.

The Face service is specifically designed for face detection and identification and it supports creating person groups and training them so the service can recognize specific people in images. It provides APIs for face detection, face verification, and identifying faces against a known set of people which is what you need to identify named staff in photos.

Custom Vision is for custom image classification and object detection and it helps you train models to classify images or detect objects but it is not specialized for managing labeled people and performing face identification of named individuals.

Form Recognizer extracts text and data from documents and scanned forms and it does not provide face detection or person identification capabilities so it is not suitable for identifying staff in photos.

Computer Vision provides general image analysis such as tagging, description, and basic object and face detection in some contexts but it does not offer the dedicated face identification and person group management features that the Face service provides. For identifying named people you should use Face instead.

Full AWS Practitioner Certification Question