Visual Search Enhancer

Understanding Metadata and AI-Powered Image Analysis in the Visual Search Enhancer

The Image Studio provides you with extensive insights into the quality of your images and their SEO parameters. Here are the various sections available.

Image Quality Score

Every asset uploaded to the Digital Asset Manager is examined and provided with an image quality score.

The image quality score factors in the following:

  • Technical Quality: Assess resolution, sharpness, exposure, color accuracy, and noise levels.
  • Aesthetic Factors: Evaluate composition, focus, lighting, and visual clarity.
  • Contextual Relevance: Ensure the image aligns with brand guidelines and project goals.
  • Metadata Completeness: Check for descriptive tags, alt text, captions, and licensing info.
  • Automation Potential: Use AI tools for scalable evaluation of sharpness, aesthetics, and subject detection.
  • Safe Search Compliance: Detect and flag inappropriate, offensive, or sensitive content.

Summary

The summary section of the image provides the following information

Properties:

  • Location: The location of the asset in the Digital Asset Library.
  • Size: The size of the asset.
  • Modified: When was the asset last modified.
  • Type: The type of asset.
  • Asset URL: A link to the asset. This can be copied and shared with users who have access to the folder.
  • Image Quality Score: The image quality score.

Metadata

The metadata section provides information that describes, identifies, and provides context for your assets, such as images, videos, documents, and other media.

The metadata types displayed here are:

Tags: Keywords or phrases used to categorize the image for easier search and retrieval.

Mark Image as Stock: Helps differentiate between proprietary and stock assets for usage rights and licensing. Also, informs search engines to not index them.

Contextual Metadata

These are descriptive fields manually added to provide context or meaning.

Title: A short title or caption identifying the image.

Description (alt): Alternative text used for accessibility or SEO, describing what’s in the image.

Embedded Metadata

This is the data automatically stored in the image file by the camera, editing software, or file processor.

AI Analysis

The Visual Search Enhancer carries out an AI Analysis of the image to determine the objects present in the image and if it is appropriate to appear in safe search on search engines.

Labels: Automatically generated keywords or categories that describe the visual content of the image. The percentage shown (e.g., Resort – 93%) indicates how confident the AI is that the label accurately describes the image. This helps classify the image for search, tagging, and categorization.

Localized Objects: Specific objects detected within defined regions of the image, along with confidence scores. This is useful for identifying distinct items in complex images (e.g., logos, buildings, products).

Safe Search: AI-generated content safety analysis based on visual cues. It assesses the image for potentially inappropriate or sensitive content. 

Categories Analyzed:

  • Adult: Nudity or explicit sexual content
  • Medical: Graphic medical procedures or imagery
  • Racy: Sexually suggestive content
  • Spoof: Fake or altered imagery (e.g., memes, manipulated photos)
  • Violence: Blood, weapons, or harmful actions

Risk Levels:

  • Very Unlikely / Unlikely: Safe for general use
  • Possible / Likely / Very Likely: Increasing levels of concern

This ensures the image complies with content safety policies, especially for public or brand-sensitive platforms.