Image descriptions, captions using azure vision

pixexid 26 Reputation points
2021-06-21T20:09:27.96+00:00

i saw a blog post https://blogs.microsoft.com/ai/azure-image-captioning/
they talking about the new image descriptions and when i tested they give me the same old descriptions107743-screen-shot-2021-06-21-at-35513-pm.png

107696-screen-shot-2021-06-21-at-35433-pm.png

how to use the new version to get the more accurate image descriptions?

Azure Computer Vision
Azure Computer Vision
An Azure artificial intelligence service that analyzes content in images and video.
371 questions
Accepted answer
  1. Stephen Howell (MSFT) 81 Reputation points
    2021-06-24T12:46:04.747+00:00

    In your code where you call the Computer Vision API, you can specify which version of the API to use.
    The latest version is 3.2, but 3.1 and 3.0 can also be used to get different results.
    For instance, API 1.0 - 3.0 don't seem to return different captions (they can return different tags though).
    3.1 and 3.2 return slightly different captions.
    If you are comfortable with Python, you can try out different API captions with this sample:
    https://github.com/stephen-howell/AI-for-Accessibility-Vision-AI-Captioning
    the line 

    analyze_url = endpoint + "vision/v" + api_version + "/analyze"

    is the line where the Azure Computer Vision endpoint is combined with the vision service, the api, and the request to analyze the results into a single URL for the REST call.

    Usage: py caption.py api_version image.png
    api_version should be 3.2 for the latest one, but try 3.1 and 3.0 to compare too.

0 additional answers

Sort by: Most helpful
Most helpful Newest Oldest

Your answer

Answers can be marked as Accepted Answers by the question author, which helps users to know the answer solved the author's problem.