In the paper, we consider the urgent need to create highly efficient hardware accelerators for machine learning algorithms, including convolutional and deep neural networks (CNN and …
We have proposed a design and simulation of hardware realizations of smart multifunctional continuous logic devices (SMCLD) as advanced basic cells of the sensor systems with …
We consider design and modeling of hardware realizations of reconfigurable multifunctional continuous logic devices (R MCL D) as advanced components of the next generation high …
VG Krasilenko, AI Nikolsky… - Independent …, 2010 - spiedigitallibrary.org
Equivalence models (EM) advantages of neural networks (NN) are shown in paper. EMs are based on vectormatrix procedures with basic operations of continuous neurologic …
In the paper we show that the biologically motivated conception of time-pulse encoding usage gives a set of advantages (single methodological basis, universality, tuning simplicity …
In the paper we show that the biologically motivated conception of the use of time-pulse encoding gives the row of advantages (single methodological basis, universality, simplicity …
In the paper, we consider the urgent need to create highly efficient hardware accelerators for machine learning algorithms, including convolutional and deep neural networks (CNN and …
In this paper, we propose a new iterative process of sorting an array of signals, which differs from the known structures of sorting signals by uniformity, versatility, which allows direct and …
In this paper, we proposed a new iterative process of sorting an array of signals, which differs from the known structures of sorting signals by uniformity, versatility, which allows …