Once your dataset has been enriched using Visual Layer’s AI models, you can begin analyzing the generated metadata to surface patterns, identify issues, and prepare the dataset for export, training, or delivery. Use the tools below to take full advantage of your enriched data.

​
Search with Semantic Models

Use Semantic Search to explore your dataset using natural language queries.

​
Common use cases:

  • Search for visual concepts (e.g., “sunset over forest”)
  • Find fine-grained differences across scenes
  • Match by meaning, not just tags

​
Filter by Issues

Detect and isolate problematic data using Visual Layer’s automated filters: These tools help you clean your dataset quickly and improve downstream performance.

​
Use Captions & Tags

Explore model-generated metadata such as:
  • Captions for entire images
  • Tags and labels for quick filtering
  • Descriptions of objects or scenes
Use Caption Search or User Tags to filter and organize your enriched data more effectively.

​
Export Enriched Data

Use the Export tool to download either the full dataset or a filtered subset.

​
Included in the export:

  • Model-generated labels
  • Captions and tags
  • Object metadata and annotations
All exported data is formatted in clean, structured JSON.