Publishing Object Detector Model in Custom Vision Service | Azure AI Solution Exam

Publishing Object Detector Model

Question

Using your custom vision service, you plan to create an object detector model.

Using the custom vision client libraries, you create a custom vision project, add tags to the project, upload the images and tag them and train the project in your active Azure subscription.

Next you plan to publish the current iteration.

Review the method call given below and complete it by choosing the appropriate parameters:

private void PublishIteration(..............................................................., Project project) 

Answers

Explanations

Click on the arrows to vote for the correct answer

A. B. C. D.

Correct Answer: C

Here is the complete method call to publish the current iteration

<pre class="brush:java;">private void PublishIteration(CustomVisionTrainingClient trainingApi, Project project)</pre>Given the completed method call in the explanation above,

Option A is incorrect.

Given the completed method call in the explanation above, Option B is incorrect.

CustomVisionPredictionClient predictionApi is used in the method call to test the current iteration.

Given the completed method call in the explanation above, Option C is correct.

For creation, training and publishing, use client CustomVisionTrainingClient.

You need to provide trainingApi for publishing the iteration.

Given the completed method call in the explanation above, Option D is incorrect.

Reference:

To learn more about object detection using custom vision, use the link given below:

The given code snippet is part of a program that creates a custom vision project and publishes its current iteration. The method PublishIteration takes two parameters: the first parameter is missing and must be chosen from the given options, and the second parameter is the custom vision project object that is already created.

To publish the current iteration of a custom vision project, we need to use the Custom Vision Training API, which provides a PublishIteration method. The CustomVisionTrainingClient class is used to call the Training API, and it provides the necessary methods to create and manage custom vision projects.

Therefore, the correct answer is option C: CustomVisionTrainingClient trainingApi. We need to use the training API to publish the iteration because it is responsible for managing the training of the custom vision project and publishing its iterations.

Option A, CustomVisionTrainingClient predictionApi, is incorrect because it uses the term "prediction" instead of "training," and prediction API is used to make predictions on new data rather than publishing the iteration of a custom vision project.

Option B, CustomVisionPredictionClient predictionApi, is also incorrect because it is used to make predictions on new data and not for publishing the iteration of a custom vision project.

Option D, CustomVisionPredictionClient trainingApi, is incorrect because it uses the term "prediction" instead of "training," and CustomVisionPredictionClient is used to make predictions on new data, not for publishing the iteration of a custom vision project.

In summary, to publish the current iteration of a custom vision project, we need to use CustomVisionTrainingClient trainingApi as the method parameter.