Step-by-Step Guide to Implementing Custom Vision in Azure | AI-102 Exam

Correct Answers:A, D, E and HOption A is CORRECT because “Install Custom Vision client library” should be the first step.

Option B is INCORRECT because “Install Custom Vision client library” should be the first step.

Option C is INCORRECT because “Get the training and prediction keys” should be the second step.

Option D is CORRECT because “Get the training and prediction keys” should be the second step.

Option E is CORRECT because “Create the Custom Vision project” should be the fifth step.

Option F is INCORRECT because “Create the Custom Vision project” should be the fifth step.

Option G is INCORRECT because “Create tags in the project” should be the sixth step.

Option H is CORRECT because “Create tags in the project” should be the sixth step.

Reference:

The correct missing steps for the image classification project with a custom vision library are:

Step 1: Create the Custom Vision project Step 2: Get the training and prediction keys Step 3: Get the sample images Step 4: Add the code Step 5: Create tags in the project Step 6: Upload and tag images Step 7: Train and publish the project Step 8: Use the prediction endpoint Step 9: Install Custom Vision client library Step 10: Run the application

Here is a detailed explanation of each step:

Step 1: Create the Custom Vision project To start, you need to create a new Custom Vision project in the Azure portal. This is where you'll train and publish your machine learning model for image classification.

Step 2: Get the training and prediction keys Next, you need to obtain your training and prediction keys for your Custom Vision project. These keys will allow you to connect to your project from your application and make predictions using the trained model.

Step 3: Get the sample images Before you can train your model, you need to have sample images to work with. You can either use your own images or download sample images from a publicly available dataset.

Step 4: Add the code Now that you have your Custom Vision project, keys, and sample images, you can start adding code to your application. You'll need to include the Azure Cognitive Services SDK and the Custom Vision client library in your project.

Step 5: Create tags in the project To train your model, you need to tag your sample images with the labels you want your model to classify. You can create tags in your Custom Vision project and then apply them to your images.

Step 6: Upload and tag images Next, you need to upload your tagged images to your Custom Vision project. You can do this using the Custom Vision portal or the Custom Vision client library.

Step 7: Train and publish the project Once you have uploaded and tagged your images, you can train your model. The training process involves using machine learning algorithms to analyze your images and create a model that can classify new images based on their tags. Once your model is trained, you can publish it to the prediction endpoint.

Step 8: Use the prediction endpoint With your model published to the prediction endpoint, you can now use it to classify new images. You can make predictions by sending HTTP requests to the endpoint, passing in the image data and your prediction key.

Step 9: Install Custom Vision client library To use the Custom Vision client library in your application, you'll need to install it using a package manager like NuGet.

Step 10: Run the application Finally, you can run your application and test it with new images to see how accurately it can classify them based on the tags you created and the model you trained.