Speech synthesis using a non-maximally decimated filter bank for embedded systems.
Nobuyuki NishizawaTsuneo KatoPublished in: INTERSPEECH (2012)
Keyphrases
- embedded systems
- filter bank
- speech synthesis
- speech recognition
- image coding
- subband
- perfect reconstruction
- low cost
- text to speech
- multiresolution
- multiscale
- computing power
- embedded devices
- embedded software
- signal processing
- frequency domain
- wavelet packet
- wavelet filters
- resource limited
- computationally efficient
- software systems
- optimal filter
- field programmable gate array
- low pass
- discrete cosine transform
- language model
- wavelet transform
- pattern recognition
- lifting scheme
- denoising
- image data
- computer vision
- artificial intelligence
- neural network