A Fully-Pipelined Hardware Design for Gaussian Mixture Models.
Conghui HeHaohuan FuCe GuoWayne LukGuangwen YangPublished in: IEEE Trans. Computers (2017)
Keyphrases
- gaussian mixture model
- hardware design
- mixture model
- em algorithm
- hardware implementation
- fpga hardware
- feature vectors
- expectation maximization
- speaker recognition
- density estimation
- probability density function
- probability density
- maximum likelihood
- speaker identification
- feature space
- covariance matrices
- field programmable gate array
- variational bayes
- machine learning
- general purpose
- embedded systems
- unsupervised learning