Design of an ultra-low Power MFCC Feature Extraction Circuit with Embedded Speech Activity Detector.
Kaiyue YangLixuan ZhuWeiwei ShanPublished in: ICTA (2021)
Keyphrases
- feature extraction
- speech recognition
- embedded systems
- feature vectors
- preprocessing
- speaker identification
- design process
- circuit design
- pattern recognition
- feature space
- activity theory
- ultra low power
- real time
- evolvable hardware
- cmos technology
- speaker recognition
- single chip
- design methodology
- speech signal
- image classification
- dimensionality reduction
- user interface
- case study