QuestionQ67

VerifiedServing and scaling models

You work for a semiconductor manufacturing company and must build a real-time application to automate the quality-control process. High-definition images of every semiconductor are captured in real time at the end of the assembly line. The photos are uploaded to a Cloud Storage bucket together with tabular data containing each semiconductor’s batch number, serial number, dimensions, and weight.

You need to configure model training and serving while maximizing model accuracy. What should you do?

Explanation

A labeled AutoML image classification model is suited to learning visual pass/fail defects from high-definition semiconductor images. Deploying the model to a Vertex AI endpoint enables online, low-latency predictions for real-time application input; batch predictions are asynchronous and intended for accumulated data when an immediate response is not required. A Pub/Sub message can notify downstream systems when the application receives a failing classification.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!