Results 271 to 280 of about 86,091 (310)
Some of the next articles are maybe not open access.
Decoupled Convolutions for CNNs
Proceedings of the AAAI Conference on Artificial Intelligence, 2018In this paper, we are interested in designing small CNNs by decoupling the convolution along the spatial and channel domains. Most existing decoupling techniques focus on approximating the filter matrix through decomposition. In contrast, we provide a two-step interpretation of the standard convolution from the filter at a single ...
Guotian Xie +4 more
openaire +1 more source
Applying CNN to Cheminformatics
2007 IEEE International Symposium on Circuits and Systems (ISCAS), 2007We describe a method for the construction of specific types of neural networks composed of structures directly linked to the structure of the molecule under consideration. Each molecule can be represented by a unique neural connectivity problem (graph) which can be programmed onto a cellular neural network.
Christian Merkwirth, Maciej Ogorzalek
openaire +1 more source
ISCAS 2001. The 2001 IEEE International Symposium on Circuits and Systems (Cat. No.01CH37196), 2002
The cellular neural network (CNN) has been widely used for associative memory. However, it has a problem called indeterminate cell. We describe this problem and propose the variable neighborhood CNN. As a result, we have been able to avoid the problem, and construct a more efficient CNN system for associative memory in simulation.
Michihiro Namba +3 more
openaire +1 more source
The cellular neural network (CNN) has been widely used for associative memory. However, it has a problem called indeterminate cell. We describe this problem and propose the variable neighborhood CNN. As a result, we have been able to avoid the problem, and construct a more efficient CNN system for associative memory in simulation.
Michihiro Namba +3 more
openaire +1 more source
International Journal of Bifurcation and Chaos, 2003
A systematic design methodology for finding CNN parameters with prescribed functions is proposed. A given function (task) is translated into several local operations, and they are realized as stable states of the CNN system. Many CNN parameters (CNN genes) with the same functions can be easily derived by using this design methodology.
Makoto Itoh, Leon O. Chua
openaire +1 more source
A systematic design methodology for finding CNN parameters with prescribed functions is proposed. A given function (task) is translated into several local operations, and they are realized as stable states of the CNN system. Many CNN parameters (CNN genes) with the same functions can be easily derived by using this design methodology.
Makoto Itoh, Leon O. Chua
openaire +1 more source
International Journal of Bifurcation and Chaos, 1997
CNN is an acronym for either Cellular Neural Network when used in the context of brain science, or Cellular Nonlinear Network when used in the context of coupled dynamical systems. A CNN is defined by two mathematical constructs: 1. A spatially discrete collection of continuous nonlinear dynamical systems called cells, where information can be ...
openaire +2 more sources
CNN is an acronym for either Cellular Neural Network when used in the context of brain science, or Cellular Nonlinear Network when used in the context of coupled dynamical systems. A CNN is defined by two mathematical constructs: 1. A spatially discrete collection of continuous nonlinear dynamical systems called cells, where information can be ...
openaire +2 more sources
2018
The main aim of this work is to present a Quaternion Phase Convolutional Neural Network. We encode 3 quaternion phases and its magnitude as an input. Our approach is bio-inspired and is expressed in one mathematical framework. The main result is to obtain a new space feature representation for deep learning which can capture non-trivial equivariant ...
Eduardo Ulises Moya-Sánchez +3 more
openaire +1 more source
The main aim of this work is to present a Quaternion Phase Convolutional Neural Network. We encode 3 quaternion phases and its magnitude as an input. Our approach is bio-inspired and is expressed in one mathematical framework. The main result is to obtain a new space feature representation for deep learning which can capture non-trivial equivariant ...
Eduardo Ulises Moya-Sánchez +3 more
openaire +1 more source
2016
Global convolutional neural networks (CNNs) activations lack geometric invariance, and in order to address this problem, Gong et al. proposed multi-scale orderless pooling(MOP-CNN), which extracts CNN activations for local patches at multiple scale levels, and performs orderless VLAD pooling to extract features.
Dan Yu, Xiao-Jun Wu 0001
openaire +1 more source
Global convolutional neural networks (CNNs) activations lack geometric invariance, and in order to address this problem, Gong et al. proposed multi-scale orderless pooling(MOP-CNN), which extracts CNN activations for local patches at multiple scale levels, and performs orderless VLAD pooling to extract features.
Dan Yu, Xiao-Jun Wu 0001
openaire +1 more source
Assamese document classification using CNN, multi-channel CNN and CNN-SVM
AIP Conference Proceedings, 2023Chayanika Talukdar, Shikhar Kumar Sarma
openaire +1 more source
CTMLP: Can MLPs replace CNNs or transformers for COVID-19 diagnosis?
Computers in Biology and Medicine, 2023Junding Sun +2 more
exaly