CLOUD-DIGITAL-LEADER practice question 12 of 35

A manufacturing organization has a large collection of images labeled as intact or defective parts. They want to use this data to build a simple solution to detect faulty parts on their production line. They have no data science expertise. Which solution should they use?

Full set: 411 questions
Answer options
A: Discovery Al for Retail
B: Pre-trainedAPIs
C: Document Al
D: AutoML Correct answer
Correct Answer:
D. AutoML
Explanation

Google Cloud AutoML Vision lets non-experts upload their own labeled images (e.g., “intact” vs “defective”) and automatically trains an image-classification model through a no-code or low-code UI. It is explicitly designed for custom vision tasks when the pre-trained Vision API cannot recognize the user’s proprietary categories, and it requires little or no data-science knowledge—exactly matching the manufacturer’s need for a simple faulty-part detector built from their labeled photos.

Why Incorrect

A. Discovery AI for Retail targets product search/recommendations for retail catalogs, not visual defect detection in manufacturing.
B. Pre-trained APIs return generic labels; they cannot learn the firm’s custom “defective vs intact” categories without additional training.
C. Document AI extracts structured data from documents (forms, invoices, etc.), unrelated to image-based quality control.

References

1. Google Cloud, “About AutoML Vision,” Section: Why AutoML Vision, para 1-2 (https://cloud.google.com/vision/automl/docs).

2. Google Cloud, “Choose between Vision API and AutoML Vision,” Table 1, Custom labels row (https://cloud.google.com/vision/automl/docs/before-you-begin).

3. Google Cloud, “Discovery AI for Retail Overview,” Introduction, para 1 (https://cloud.google.com/retail/docs/overview).

4. Google Cloud, “Document AI Overview,” What Document AI does, para 1 (https://cloud.google.com/document-ai/docs/overview).

Scroll to Top