A retail analytics team is scoping four candidate features for a new store-monitoring system: counting how many shoppers pass through the front entrance by detecting people in video frames, drawing a bounding box around each shopping cart in a camera frame and labeling it 'cart', condensing a week of customer email complaints into a short list of recurring themes, and predicting next month's hourly foot traffic from two years of historical hourly counts. (Select TWO.) Which two features are computer vision workloads?

The correct answers are counting shoppers passing through the entrance by detecting people in video frames, and drawing a bounding box around each shopping cart in a frame and labeling it 'cart'. Both involve locating and identifying objects within images, which is the core of a computer vision workload. Condensing a week of customer email complaints into a short list of recurring themes is incorrect because it operates on written text rather than pixels, making it a natural language processing task. Predicting next month's hourly foot traffic from two years of historical hourly counts is incorrect because it predicts a future numeric value from a historical time series, making it a forecasting task built on structured, sequential data rather than image input. The distinguishing factor across all four features is the type of data being processed and the kind of insight being produced: the two correct features both take raw video frames and locate or classify visual elements within them, while the complaint-summarization feature works on text and the foot-traffic-prediction feature works on time-series numbers, placing them in different AI workload categories entirely.