Struggling with Azure Custom Vision model retraining? Discover the optimal workflow for auto-tagging new images while preparing for the AI-102 exam. Learn the correct sequence to minimize manual labeling time and boost efficiency in Azure AI Vision projects. Essential for Azure certification candidates!
Table of Contents
Question
You are training a Custom Vision model. You have a large number of new images that you need to retrain the model. You need to generate tags for the new images. You want to minimize the time to retrain the model.
What should you do?
Place the appropriate choices in the correct order.
Unordered Choices:
- Train the model.
- Tag the images manually
- Get suggested tags.
- Confirm tags.
- Upload the new images to your Custom Vision project.
- Provide some initial training data by labeling a portion of your dataset with an equal number of images for each tag.
Answer
Correct Order:
- Upload the new images to your Custom Vision project.
- Provide some initial training data by labeling a portion of your dataset with an equal number of images for each tag.
- Train the model.
- Get suggested tags.
- Confirm tags.
Explanation
You would perform the following steps:
- Upload the new images to your Custom Vision project.
- Provide some initial training data by labeling a portion of your dataset with an equal number of images for each tag.
- Train the model.
- Get suggested tags.
- Confirm tags.
Smart Labeler is a feature in Azure Vision that is designed to lessen the time of tagging images during the training of a Custom Vision model.
You would first upload the images for training to your Custom Vision project.
You would then provide some initial training data by labeling a portion of your dataset. It is recommended that an equal number of images be chosen for each tag.
You would start training the model with the labeled data.
Once the training has finished, the Smart Labeler suggests tags for the remaining untagged images.
You would not tag the images manually. Since you want to minimize the time taken to retrain the model, using the suggested tags from the Smart Labeler will speed up tagging the images for training.
You would then confirm the tags. Suggested tags are sorted based on prediction uncertainty. You need to apply the object detection tags to the images in the gallery.
Microsoft Azure AI Engineer Associate AI-102 certification exam practice question and answer (Q&A) dump with detail explanation and reference available free, helpful to pass the Microsoft Azure AI Engineer Associate AI-102 exam and earn Microsoft Azure AI Engineer Associate AI-102 certification.