A 0.4 V 298 nJ/op Neural Signal Spectral Feature Extraction Module With Novel Approximate MACs and Custom Compressors.
Weiwei ShiYuan XuChiu-Sing ChoyZhiyong ChenJunwei YangRobert K. F. TengMei JiangX. DengPublished in: IEEE Trans. Circuits Syst. II Express Briefs (2019)
Keyphrases
- feature extraction
- continuous wavelet transform
- frequency domain
- neural classifier
- neural network
- signal analysis
- network architecture
- hilbert huang transform
- frequency spectrum
- signal processing
- image classification
- data compression
- feature selection
- application specific
- non stationary
- preprocessing
- image processing
- dimension reduction
- pattern classification
- support vector machine svm
- linear feature extraction
- spectral estimation
- feature space
- wavelet transform
- high frequency
- domain specific
- extracted features
- short time fourier transform
- discriminant analysis
- face recognition
- feature vectors
- spectral analysis
- biologically inspired
- principal component analysis
- spatial frequency
- hyperspectral
- signal detection
- original signal
- texture analysis
- hyperspectral images
- multispectral images