The speech recognition process is representable as a two-level hierarchical process, comprise of a low level sub words recognition, for example, phoneme recognition, along with high-level—words, sentences, contextual information recognition, language analysis; every one of the levels being representable in a recursive manner as numerous other amounts of processing.
- Hybrid Systems, neural network lecture notes
- Time-Delay Neural Networks for Speech Recognition, neural network lecture notes
- MLP for Speech Recognition, classroom notes, downloads
- Hybrid systems for speech recognition and language processing free ebook download pdf
- Speech Recognition Using Wavelet Transform