Spoken word recognition based on top-down phoneme segmentation.
Kiyoaki AikawaMasahide SugiyamaKiyohiro ShikanoPublished in: ICASSP (1985)
Keyphrases
- segmentation algorithm
- image segmentation
- correct recognition
- image parsing
- recognition accuracy
- recognition rate
- numeral strings
- object recognition
- symbol recognition
- multiscale
- image analysis
- level set
- medical images
- feature extraction and matching
- automatic speech recognition systems
- skeleton extraction
- features extraction
- segmentation accuracy
- automatic recognition
- recognition algorithm
- segmentation method
- speech recognition
- edge detection
- pattern recognition
- figure ground
- shape prior
- figure ground segmentation
- fully automatic
- hidden markov models
- sign recognition
- phoneme recognition
- context dependent
- object segmentation
- fully unsupervised