M5HisDoc: A Large-scale Multi-style Chinese Historical Document Analysis Benchmark.
Yongxin ShiChongyu LiuDezhi PengCheng JianJiarong HuangLianwen JinPublished in: NeurIPS (2023)
Keyphrases
- document analysis
- word segmentation
- document images
- image analysis
- character recognition
- document image analysis
- english text
- document processing
- word recognition
- real world
- text analysis
- handwriting recognition
- word level
- printed documents
- historical documents
- document image retrieval
- document layout
- electronic documents
- video analysis
- text mining
- multimedia