Which description best matches the computer vision family of AI workloads?
Show answer and explanation
Answer: B
Computer vision processes pixel data to derive information about the visual world: what objects are present, where they are, what text appears in a scene, or whether an image looks normal. Typical tasks include classification, object detection, segmentation and optical character recognition. Modern services expose these as prebuilt or custom models, and the workload is distinguished by its visual input rather than by the algorithm used.
Why the other options are wrong
- A. Written documents are handled by text and language workloads; reading a scan as an image is vision, but understanding the words is language processing.
- C. Financial market projection is a forecasting workload built on numeric time series.
- D. Spoken conversation is handled by speech and natural language processing, even though spectrogram images can be used internally.
Exam tip. Pictures and video point to computer vision.