QuTI! Quantifying Text-Image Consistency in Multimodal Documents.
Matthias SpringsteinEric Müller-BudackRalph EwerthPublished in: CoRR (2021)
Keyphrases
- image data
- information retrieval
- image analysis
- image content
- free text
- multiscale
- keywords
- image features
- scanned documents
- text information
- textual information
- single image
- text retrieval
- image retrieval
- input image
- web images
- web documents
- image classification
- edge detection
- high resolution
- multi modal
- image representation
- plagiarism detection
- textual content
- document analysis
- printed documents
- text documents
- image collections
- latent semantic analysis
- image segmentation
- textual descriptions
- document level
- document processing
- textual data
- key concepts
- text queries
- scanned images
- digital libraries
- similarity measure
- electronic documents
- digital documents
- page layout
- historical manuscripts
- scanned document images
- handwritten documents
- document categorization
- multimedia documents
- text collections
- relevant documents
- bag of words
- semantic information
- document collections
- information retrieval systems
- xml documents
- metadata