Convert Standard Domain to Compact Domain for Real-time Classification | AI-102 Exam Preparation

Convert Standard Domain to Compact Domain

Question

The capability to export classifiers is offered in the Custom Vision Service.

You may use this feature in your application locally for real-time classifications.

Since Custom Vision Service only exports compact domains, you plan to convert standard domain to compact domain to run the classifier locally for real-time classification.

Review the scenario given above and select the answer choice where the sequence of steps to convert standard to a custom domain iscorrectly maintained:

Answers

Explanations

Click on the arrows to vote for the correct answer

A. B. C. D.

Correct Answer: D.

Option A is incorrect because you would need to select the project first and change it to the compact domain.

The next step would be to retain the model and export it in the appropriate format.

Option B is incorrect because you would need to retain the model before exporting it in the appropriate format.

Option C is incorrect because you would need to select the project first and change it to the compact domain.

The next step would be to retain the model and export it in the appropriate format.

Option D is correct because following this sequence, you can export the model to execute locally for real-time classification.

To learn more about Custom Vision Service export options, use the link given below:

The correct sequence of steps to convert a standard domain to a compact domain in Custom Vision Service and export the model is:

C. Select & save the new domain -> Select the project -> Retrain the model -> Export the model

Here is a detailed explanation of each step:

  1. Select & save the new domain: In Custom Vision Service, a domain is a pre-defined set of labels and image types that the classifier can recognize. To convert a standard domain to a compact domain, you need to create a new domain and choose which labels and image types to include. This step involves selecting and saving the new domain.

  2. Select the project: A project in Custom Vision Service is a collection of images and labels that you use to train a classifier. Once you have created the new domain, you need to select the project that you want to use to train the classifier.

  3. Retrain the model: After you have selected the project, you need to retrain the classifier using the new domain. This step involves uploading images, labeling them, and training the classifier to recognize the labels in the new domain. During the training process, you can adjust the model parameters to improve its accuracy.

  4. Export the model: Once the model has been trained on the new domain, you can export it to use it locally for real-time classifications. The exported model will be in the form of a compact domain, which is optimized for real-time inference.

Option A is incorrect because it retrained the model before selecting the project. The correct sequence requires selecting the project before retraining the model.

Option B is incorrect because it exports the model before retraining the model. The correct sequence requires retraining the model before exporting it.

Option D is incorrect because it exports the model before retraining the model. The correct sequence requires retraining the model before exporting it, as explained above.