BN-HTRd: A Benchmark Dataset for Document Level Offline Bangla Handwritten Text Recognition (HTR) and Line Segmentation.
Md. Ataur RahmanNazifa TabassumMitu PaulRiya PalMohammad Khairul IslamPublished in: CoRR (2022)
Keyphrases
- benchmark datasets
- document level
- handwritten text recognition
- sentence level
- language model
- bayesian networks
- image segmentation
- sentiment classification
- hidden markov models
- handwritten documents
- level set
- document retrieval
- probabilistic model
- pseudo relevance feedback
- text lines
- statistical machine translation