nerdexam
Google

PROFESSIONAL-MACHINE-LEARNING-ENGINEER · Question #215

You work for an auto insurance company. You are preparing a proof-of-concept ML application that uses images of damaged vehicles to infer damaged parts. Your team has assembled a set of annotated…

The correct answer is B. Train an object detection model in AutoML by using the annotated image data. To quickly create an initial object detection model for inferring damaged car parts from annotated images with bounding boxes, the most efficient approach is to use AutoML Vision's object detection capability.

Submitted by skyler.x· Apr 18, 2026ML model development

Question

You work for an auto insurance company. You are preparing a proof-of-concept ML application that uses images of damaged vehicles to infer damaged parts. Your team has assembled a set of annotated images from damage claim documents in the company's database. The annotations associated with each image consist of a bounding box for each identified damaged part and the part name. You have been given a sufficient budget to train models on Google Cloud. You need to quickly create an initial model. What should you do?

Options

  • ADownload a pre-trained object detection model from TensorFlow Hub. Fine-tune the model in
  • BTrain an object detection model in AutoML by using the annotated image data.
  • CCreate a pipeline in Vertex AI Pipelines and configure the AutoMLTrainingJobRunOp component
  • DTrain an object detection model in Vertex AI custom training by using the annotated image data.

How the community answered

(48 responses)
  • A
    17% (8)
  • B
    71% (34)
  • C
    8% (4)
  • D
    4% (2)

Why each option

To quickly create an initial object detection model for inferring damaged car parts from annotated images with bounding boxes, the most efficient approach is to use AutoML Vision's object detection capability.

ADownload a pre-trained object detection model from TensorFlow Hub. Fine-tune the model in

Fine-tuning a pre-trained model requires more hands-on ML expertise, model selection, and coding compared to AutoML, making it less suitable for quickly creating an initial model with minimal effort.

BTrain an object detection model in AutoML by using the annotated image data.Correct

AutoML Vision Object Detection is specifically designed for situations where annotated image data (with bounding boxes and labels) is available, enabling users to quickly train high-quality object detection models with minimal ML expertise and coding, which perfectly fits the 'quickly create an initial model' requirement for a proof-of-concept.

CCreate a pipeline in Vertex AI Pipelines and configure the AutoMLTrainingJobRunOp component

While using Vertex AI Pipelines to trigger an AutoML training job is good for MLOps, it adds an unnecessary layer of complexity for the initial goal of *quickly creating an initial model* for a PoC.

DTrain an object detection model in Vertex AI custom training by using the annotated image data.

Vertex AI custom training offers the most flexibility but requires significant ML expertise to develop, train, and optimize a model, making it a higher effort and slower approach for quickly building an initial model compared to AutoML.

Concept tested: Rapid object detection model development with AutoML

Source: https://cloud.google.com/vertex-ai/docs/vision/object-detection/overview

Topics

#AutoML#Object Detection#Rapid Prototyping#Vertex AI

Community Discussion

No community discussion yet for this question.

Sign up to join the discussion
Full PROFESSIONAL-MACHINE-LEARNING-ENGINEER Practice