According to Microsoft's AI-900 study guide and the Microsoft Learn module "Identify common types of AI workloads", computer vision is the branch of AI that enables machines to interpret and act upon visual inputs such as photos, videos, or camera feeds.
In this scenario, the recycling machine must identify bottles based on their shape and reject other objects. This requires the AI system to visually analyze an image or a video stream and make a decision based on shape recognition-a classic computer vision application.
Computer vision models are trained with labeled image data (e.g., images of correct bottle shapes vs. incorrect items). Using techniques such as object detection or image classification, the model learns to detect patterns and identify whether an image matches the desired object.
The other options are not suitable:
* A. Anomaly detection identifies unusual data patterns, not visual features.
* B. Conversational AI handles human-computer interactions via text or voice.
* D. Natural language processing (NLP) deals with understanding and generating human language.
Therefore, because the task requires visual recognition of bottle shapes, the appropriate AI workload is Computer Vision.
Reference:Microsoft Learn - Identify common types of AI workloads: Computer Vision, NLP, Conversational AI, and Anomaly Detection (AI-900 Learning Path)