Cepstral and acoustic ternary pattern based hybrid feature extraction approach for end-to-end bangla speech recognition.
Mohit Dua AkankshaShelza DuaPublished in: J. Ambient Intell. Humaniz. Comput. (2023)
Keyphrases
- speech recognition
- end to end
- feature extraction
- speech recognition systems
- speaker independent
- speech recognizers
- mel frequency cepstral coefficients
- speaker identification
- cepstral coefficients
- pattern recognition
- acoustic models
- speech signal
- formant frequencies
- automatic speech recognition
- hidden markov models
- noisy speech
- speech recognizer
- noisy environments
- language model
- speech synthesis
- speech recognition technology
- face recognition
- speech retrieval
- wavelet transform
- acoustic features
- feature space
- feature selection
- speaker recognition
- congestion control
- handwriting recognition
- frequency domain
- principal component analysis
- feature vectors
- language identification
- speech enhancement