FPGA-Based Low-Power Speech Recognition with Recurrent Neural Networks.
Minjae LeeKyuyeon HwangJinhwan ParkSungwook ChoiSungho ShinWonyong SungPublished in: CoRR (2016)
Keyphrases
- low power
- recurrent neural networks
- speech recognition
- high speed
- power consumption
- low cost
- neural network
- language model
- hidden markov models
- feed forward
- automatic speech recognition
- recurrent networks
- speech processing
- pattern recognition
- speech recognizer
- speech synthesis
- application specific
- logic circuits
- noisy environments
- hardware implementation
- low power consumption
- artificial neural networks
- real time
- echo state networks
- speaker identification
- speech signal
- speech recognition systems
- speaker independent
- speech recognizers
- isolated word
- gate array
- image processing
- field programmable gate array
- ultra low power