
Ultimate access to all questions.
You work for a manufacturing company that sources 750 different components from 750 suppliers. Your team has a labeled dataset with approximately 1000 examples of each unique component. The goal is to develop an app within a few working days that can help warehouse workers identify incoming components by analyzing photos of them. What is the most efficient approach to create a Proof-Of-Concept for this app?
A
Train your own image recognition model using transfer learning techniques.
B
Use Cloud Vision API and provide custom labels as hints for recognition.
C
Use Cloud Vision AutoML, but reduce the size of the dataset by half.
D
Use Cloud Vision AutoML with the existing dataset.