MSP-MFCC: Energy-Efficient MFCC Feature Extraction Method With Mixed-Signal Processing Architecture for Wearable Speech Recognition Applications.
Qin LiYuze YangTianxiang LanHuifeng ZhuQi WeiFei QiaoXinjun LiuHuazhong YangPublished in: IEEE Access (2020)
Keyphrases
- speech recognition
- energy efficient
- signal processing
- sensor networks
- cepstral coefficients
- wireless sensor networks
- speech processing
- speech signal
- pattern recognition
- energy consumption
- language model
- multi core architecture
- hidden markov models
- automatic speech recognition
- energy efficiency
- speech recognizer
- base station
- image processing
- speaker identification
- speech synthesis
- noisy environments
- speech recognition technology
- mel frequency cepstral coefficients
- speaker diarization
- speech recognition systems
- routing protocol
- feature vectors
- speaker independent
- noisy speech
- routing algorithm
- feature set
- sensor nodes
- sensor data
- data transmission
- background noise
- shortest path
- speaker adaptation
- artificial intelligence
- feature selection