What This Section Covers

Get a high-level understanding of the tools available in Visual Layer for search, filtering, quality control, and data curation. Use this section as your launchpad for working with large visual datasets.
This Section Covers:
  • Semantic Search: Locate relevant data using natural language-powered search across your dataset.
  • Caption search: Surface relevant content by matching queries to labels and captions.
  • Visual Search Find similar images and objects through visual comparison.
  • Filter Dataset Narrow down data based on specific attributes, labels, or conditions.
  • Detect Mislabeled Identify and correct incorrectly labeled data to improve dataset accuracy.
  • Detect Unlabeled Find and review entries that are missing annotations.
  • Detect Duplicates Spot and manage duplicate entries in your dataset.
  • Detect Outliers Detect irregular data points that may affect analysis or model performance.
  • Select Unique Items Extract only essential samples for AI model training.
  • Dataset quality Detect and manage image issues (e.g., blurry, dark, or overexposed) to ensure clean, usable data.