GeoLayoutLM: Geometric Pre-training for Visual Information Extraction.
Chuwei LuoChangxu ChengQi ZhengCong YaoPublished in: CVPR (2023)
Keyphrases
- information extraction
- text mining
- precision and recall
- visual features
- visual information
- machine learning
- training process
- training algorithm
- natural language processing
- test set
- named entities
- geometric information
- machine translation
- question answering
- training samples
- training examples
- web mining
- knowledge discovery
- word sense disambiguation
- free text
- named entity recognition
- natural language
- image sequences
- geometric structure
- visual cues
- training phase
- information retrieval