LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking.
Yupan HuangTengchao LvLei CuiYutong LuFuru WeiPublished in: ACM Multimedia (2022)
Keyphrases
- single image
- image data
- image classification
- input image
- image retrieval
- scanned documents
- web documents
- image features
- multiscale
- image analysis
- image content
- information retrieval
- image segmentation
- web images
- keywords
- printed documents
- text retrieval
- information retrieval systems
- text information
- text documents
- expert systems
- document images
- document retrieval
- edge detection
- semantic labels
- training set
- document processing
- text lines
- text content
- empirically derived
- handwritten documents
- document analysis
- multimedia documents
- image representation
- low level
- high resolution
- artificial intelligence