Using Various Features in Machine Learning to Obtain High Levels of Performance for Recognition of Japanese Notational Variants.
Masahiro KojimaMasaki MurataJun'ichi KazamaKow KurodaAtsushi FujitaEiji AramakiMasaaki TsuchidaYasuhiko WatanabeKentaro TorisawaPublished in: PACLIC (2010)
Keyphrases
- high levels
- machine learning
- feature extraction
- pattern recognition
- machine learning approaches
- extracting features
- feature extractor
- image features
- feature space
- object recognition
- reinforcement learning
- object class recognition
- robust recognition
- automatic recognition
- feature engineering
- recognition rate
- discrimination power
- visual object recognition
- learning algorithm
- feature set
- low level
- machine learning algorithms
- supervised learning
- video sequences
- feature vectors
- hidden markov models
- information extraction
- handwritten characters
- feature generation
- feature analysis
- text mining
- co occurrence
- character recognition
- recognition accuracy
- machine learning methods