AWS Certified AI Practitioner Practice Exam

What strategy should an AI practitioner use to automatically categorize animals in photos?

1. A. Object detection

2. B. Anomaly detection

3. C. Named entity recognition

4. D. Inpainting

The correct answer is: A. Object detection

Using object detection is the most appropriate strategy for automatically categorizing animals in photos. Object detection algorithms are specifically designed to locate and identify objects within images by bounding them with boxes and assigning labels. This technique is effective for tasks where multiple categories exist in a single image, allowing the application to recognize various animals present and classify them accordingly. Object detection works by training a model on annotated images containing different animals, where each animal is labeled with a specific class name. Once trained, the model can accurately predict and classify animals in new, unseen photos by detecting the objects represented within them. This capability is essential in scenarios such as wildlife conservation, where identifying species from images can help researchers monitor animal populations. The other strategies listed are not suited for directly categorizing animals in photos. Anomaly detection focuses on identifying unusual patterns or outliers in data rather than categorization. Named entity recognition is primarily used in natural language processing to identify and categorize entities from text, not visual data. Inpainting refers to the process of reconstructing missing parts of an image, which does not relate directly to categorizing objects within a photo. Thus, object detection stands out as the most effective choice for this purpose.