Question

A retail company wants to build a lightweight vision app in Azure AI Foundry. When a merchandiser uploads a product photo, the app should return a short, human-readable description of what is shown in the image so it can be used as accessibility alt-text on the website. Which approach should the developer take in Foundry to implement this feature?

Accepted Answer

Deploy a multimodal generative model and send the uploaded image with a prompt asking it to describe the product shown.

Answer

Deploy a text-only chat model and pass the image file path as a string in the user prompt.

Answer

Use an image generation model to create a new image and return its file name as the description.

Answer

Use Azure Content Understanding to redact personally identifiable information from the photo.