Data Types

DocArray lets you represent text, image, video, audio, and 3D meshes as Documents, whether separate, nested or combined, and process them as a DocumentArray. Here are some motivating examples:

• Text
  • Segment long Documents
  • Convert text to ndarray
  • Convert ndarray back to text
  • Simple text matching with feature hashing
  • Searching at chunk level with subindex
• Image
  • Load image data
  • Simple image processing
  • Display image sprite
  • Segment large complicated image into small ones
• Video
  • Load video data
  • Key frame extraction
  • Save as video file
  • Create Document from webcam
• Audio
  • Load .wav file
  • Save as .wav file
  • Example
  • Other tools & libraries for audio data
• 3D Mesh
  • Vertices and faces representation
  • Point cloud representation
  • RGB-D image representation
• Table
  • Load CSV table
  • Save as CSV file
• Multi-modal
  • Model your data
    • Instantiate a Document
    • Nested dataclasses and list types
  • Access the data
  • Generate embeddings
    • Embed each modality
    • Embed parent Document
  • Perform search
    • Find Document by modality embedding
    • Find Document by combined embedding