A multi-granularity contrastive framework, Hi-GITA, aligns Chinese character images with stroke, radical, and structure sequences and improves zero-shot recognition accuracy by large margins on several benchmarks.
M5hisdoc: A large-scale multi-style chinese historical document analysis benchmark
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Zero-Shot Chinese Character Recognition with Hierarchical Multi-Granularity Image-Text Aligning
A multi-granularity contrastive framework, Hi-GITA, aligns Chinese character images with stroke, radical, and structure sequences and improves zero-shot recognition accuracy by large margins on several benchmarks.