CICA: Content-Injected Contrastive Alignment for Zero-Shot Document Image Classification.
Sankalp SinhaMuhammad Saif Ullah KhanTalha Uddin SheikhDidier StrickerMuhammad Zeshan AfzalPublished in: CoRR (2024)
Keyphrases
- image classification
- textual content
- document content
- relevant content
- web documents
- multimedia documents
- bag of words
- structured documents
- content and structure
- feature extraction
- retrieval systems
- document representation
- image alignment
- effective retrieval
- image representation
- text content
- keywords
- semi structured documents
- document structure
- pdf files
- metadata
- document classification
- semantic information
- image features
- dynamic time warping
- visual words
- content similarity
- document images
- visual features
- electronic documents
- information retrieval
- multimedia
- scientific papers
- feature space
- word level
- multi label
- text documents
- document retrieval
- object categories