Gibbs sampling with low-power spiking digital neurons.
Srinjoy DasBruno Umbria PedroniPaul MerollaJohn V. ArthurAndrew S. CassidyBryan L. JacksonDharmendra S. ModhaGert CauwenberghsKenneth Kreutz-DelgadoPublished in: ISCAS (2015)
Keyphrases
- low power
- gibbs sampling
- mixed signal
- neuron model
- high speed
- low cost
- power consumption
- single neuron
- spiking neural networks
- hodgkin huxley
- spike trains
- spiking neurons
- markov chain
- parameter estimation
- topic models
- expectation maximization
- latent dirichlet allocation
- approximate inference
- em algorithm
- belief networks
- multi channel
- neural network
- markov chain monte carlo
- low power consumption
- graphical models
- image sensor
- ultra low power
- posterior distribution
- bayesian inference
- model selection
- maximum likelihood
- co occurrence
- state space
- probabilistic model
- video sequences