IDocumentClassifier<T>

Interfaces · AiDotNet.Document.Interfaces

Interface for document classification models.

Implements: IDocumentModel<T>

Known implementations: DiT<T>, DocFormer<T>, UDOP<T>

For Beginners

Document classification is like sorting mail into different piles. The model looks at a document and decides what type it is. This is useful for:

  • Organizing scanned documents
  • Routing documents to appropriate processing pipelines
  • Quality control in document processing

Example usage:

How It Works

Document classification models categorize documents into predefined classes such as invoices, forms, letters, scientific papers, etc.

Properties

Property Summary
IReadOnlyList<string> AvailableCategories { get; } Gets the available classification categories for this model.

Methods

Method Summary
DocumentClassificationResult<T> ClassifyDocument(Tensor<T> documentImage) Classifies a document image into predefined categories.
DocumentClassificationResult<T> ClassifyDocument(Tensor<T> documentImage, int topK) Classifies a document and returns top-K predictions.