Unsupervised font reconstruction based on token co-occurrence.
Michael Patrick CutterJoost van BeusekomFaisal ShafaitThomas M. BreuelPublished in: ACM Symposium on Document Engineering (2010)
Keyphrases
- co occurrence
- unsupervised learning
- texture analysis
- semantic similarity
- wordnet
- named entities
- latent semantic analysis
- semantic relations
- semantic relationships
- semi supervised
- topic models
- occurrence frequency
- artificial intelligence
- word co occurrence
- unsupervised manner
- visual words
- object detection
- probabilistic model
- pairwise