Automatically Populate Image Properties with AI

The Smart Image Analyzer uses advanced AI capabilities to analyze images and automatically populate media object properties. When images are uploaded to Optimizely CMS, the feature extracts valuable information and fills in key attributes seamlessly. Here are some examples of what it can do:

1. Alt Text for Accessibility

   • Alt text is automatically created in multiple languages, enhancing accessibility and SEO.

2. Extract Text from Images

   • Using OCR (Optical Character Recognition), the analyzer identifies and captures text within images, storing it in a semicolon-separated format.

3. Generate Tags for Improved Searchability

   • Tags are automatically generated based on objects and elements detected in the image, presented as a comma-separated list.

4. Create Persuasive Sales Descriptions

   • The AI crafts product-oriented descriptions, optimized for sales pitches, leaving out unnecessary details to keep it concise and impactful.

5. Identify Objects and Count People

   • For specialized needs, the analyzer can detect specific objects (e.g., cars) or count the number of people in an image, providing Boolean or numeric outputs.

How It Works

The Smart Image Analyzer leverages predefined attributes set in your Optimizely codebase on Image Model. Here are a few example properties:

• • [AnalyzeImageAltText]
    Produces Alt text for enhanced localization.

• [AnalyzeImageAltText(languageCode: "sv")]
  Produces Swedish Alt text for enhanced localization.

• [AnalyzeImage(prompt: "Return tags from the image on objects in the picture. Return a comma-separated list.")]
  Automatically generates tags based on image content.
•  

• [AnalyzeImage(prompt: "How many people are in the picture? Return an integer.")]
  Counts the number of people visible in the image.

This automated approach ensures consistency and saves significant time for editors, freeing them to focus on crafting compelling content rather than managing metadata.

Benefits for Editors

1. Enhanced Productivity

   • Automating image analysis reduces manual tasks, speeding up workflows and allowing editors to allocate time to higher-value activities.

2. Improved Accuracy

   • AI-driven metadata generation minimizes human error and enhances consistency across your media library.

3. SEO and Accessibility Gains

   • Automatically generated alt text and tags optimize images for search engines and ensure compliance with accessibility standards.

4. Multilingual Support

   • Localization features make it easier to tailor content for global audiences.

Ready to Transform Your Workflow?

The Smart Image Analyzer is a revolutionary tool for content editors using Optimizely CMS. By leveraging AI to automate image management, this tool not only improves efficiency but also enhances the quality of your content. 

Empower your content strategy with AI-driven image management today!

Documentation

Find our documentation on github: https://github.com/Epicweb-Optimizely/Epicweb.Optimizely.AIAssistant/blob/master/configuration-image-analyzer.md

Code samples:

The following are examples of properties that can be used with AnalyzeImageAttribute:

1. Alt Text in Multiple Languages:

    [AnalyzeImageAltText]
    public virtual IList<LocalizedString> AltTextList { get; set; }
   

   

2. Alt Text in one specific language:

   [AnalyzeImageAltText(languageCode: "sv")]
   public virtual string AltTextSE { get; set; }

   

3. Generate Tags:

   [AnalyzeImage(prompt: "Return tags from the image on objects in the picture. Return a comma-separated list.")]
   public virtual IList<string> Tags { get; set; }

   

4. Generate Sales Description:

   [AnalyzeImage(prompt: "Describe this as a persuasive sales pitch, you are selling a product.", languageCode: "en")]
   public virtual string SalesDescription { get; set; }

   

5. Identify Objects:

    [AnalyzeImage(prompt: "Return true or false if the image contains one or more cars. Return true or false only.")]
    public virtual bool IsCarInImage { get; set; }

6. Count People:

    [AnalyzeImage(prompt: "How many people is there in the picture? return an int")]
    public virtual int PeopleCountTest { get; set; }

7. Extract Text from Images OCR (Optical Character Recognition):

    [AnalyzeImage(prompt: "Return the Text that appear in the image. Return in semi colon separated (;) text.")]
    public virtual string TextInImage { get; set; }