Improved Recognition of Spontaneous Hungarian Speech - Morphological and Acoustic Modeling Techniques for a Less Resourced Task.
Péter MihajlikZoltán TüskeBalázs TarjánBottyán NémethTibor FegyóPublished in: IEEE Trans. Speech Audio Process. (2010)
Keyphrases
- speech sounds
- speech recognition systems
- recognition engine
- speech recognition
- speaker independent
- pattern recognition
- text to speech
- recognition rate
- feature extraction
- object recognition
- automatic speech recognition systems
- noisy speech
- visual speech
- source localization
- speech synthesis
- recognition process
- automatic speech recognition
- recognition algorithm
- speech signal
- noisy environments
- image processing
- speech corpus
- recognition accuracy
- word recognition
- character recognition
- text recognition
- spoken language
- binary images
- facial action units
- action recognition
- hidden markov models
- dialogue system