Google Professional Machine Learning Engineer
Get started today
Ultimate access to all questions.
Ultimate access to all questions.
You are working at a semiconductor manufacturing company and tasked with developing a real-time application to automate the quality control process. At the end of the assembly line, high-definition images of each semiconductor are captured in real time. These photos, along with accompanying tabular data containing each semiconductor’s batch number, serial number, dimensions, and weight, are uploaded to a Cloud Storage bucket. Your goal is to configure the model training and serving setup to maximize model accuracy. What should you do?
Explanation:
Option A is the correct answer. This approach involves using Vertex AI Data Labeling Service to label the images, which is crucial for image classification tasks. An AutoML image classification model is then trained to distinguish between passing and failing semiconductors. Deploying the model in real-time ensures that quality control decisions can be made promptly. Configuring Pub/Sub to publish a message when an image is categorized into the failing class allows for immediate action to be taken on defective semiconductors, streamlining the quality control process effectively.