Quantum Multilayer Self Organizing Neural Network (QMLSONN) Architecture

Zi Ye, Kai Yu, Gongde Guo +1 more

Siddhartha Bhattacharyya, Pankaj Pal, Sandip Bhowmick

Binary object extraction from a noisy perspective is a challenging proposition in the computer vision community. Various research initiatives based on different methodologies have been entrusted on this aspect over the decades. The multilayer self-organizing neural network (MLSONN) architecture implemented by a fuzzy measure guided backpropagation of errors is one of the best accomplishments in this direction. In this article, a quantum multilayer self-organizing neural network (QMLSONN) architecture is proposed to achieve the same objective. Quantum computation plays a remarkable role in the proposed architecture. The proposed architecture operates using single qubit rotation gates. Here different nodes and interconnection weights act as vital components for this intention. The proposed QMLSONN architecture comprises three processing layers specified as input, hidden and output layers. The nodes in the processing layers are represented by qubits and the interconnection weights are represented by quantum gates. A quantum measurement at the output layer destroys the quantum states of the processed information thereby inducing assimilation of linear indices of fuzziness as the network system errors used to adjust network interconnection weights through a proposed quantum back propagation algorithm. Results of application of the QMLSONN are established on a synthetic and a real life spanner image with various degrees of Gaussian noise and uniform noise. Comparative study with the classical MLSONN architecture reveals the time efficiency of the presented QMLSONN architecture. Moreover, the QMLSONN architecture also restores the shape of the extracted objects to a great extent.

Jesse Van Griensven Thé, Roydon Andrew Fraser, José R. Rosas-Bustos

Ajit Narayanan, Tammy Menneer

Siddhartha Bhattacharyya, Pankaj Pal, Sandip Bhowmik

Proper extraction of objects from a noisy perspective is an upheaval task in the computer vision research community. Several intelligent research paradigms have been focused on this aspect over the years. Notable among them is the multilayer self organizing neural network (MLSONN) architecture assisted by fuzzy measure guided back propagation of errors. In this article, we propose a quantum version of the MLSONN architecture which operates using single qubit rotation gates. The proposed QMLSONN architecture comprises three processing layers viz., input, hidden and output layers. The nodes of the processing layers are represented by qubits and the interconnection weights are represented by quantum gates. A quantum measurement at the output layer destroys the quantum states of the processed information thereby inducing incorporation of linear indices of fuzziness as the network system errors used to adjust network interconnection weights through a proposed quantum back propagation algorithm. Results of application of the QMLSONN are demonstrated on a synthetic and a real life spanner image with various degrees of Gaussian noise. A comparative study with the performance of the classical MLSONN architecture reveals the time efficiency of the proposed QMLSONN architecture.

Siddhartha Bhattacharyya, Pankaj Pal, Sandip Bhowmick

In this article, a quantum version of the parallel self organizing neural network (QPSONN) architecture for extraction of pure color objects from a noisy perspective is proposed. The QPSONN architecture operates in a phased manner to process input noisy pure color images. After the segregation of the pure color inputs into pure color components in the initial phase, these components are subsequently forwarded for processing to three component quantum multilayer self organizing neural network (QMLSONN) architectures composed of three processing layers viz., input, hidden and output layers characterized by qubits based neurons. The interconnection weights are represented by single qub it rotation gates. Quantum measurements at the component output layers destroy the quantum states of the processed information facilitating adjustment of network interconnection weights by a quantum back propagation algorithm using linear indices of fuzziness. Finally, a fusion of the stable component outputs are brought about in a sink layer to produce extracted outputs. Results of application of the QPSONN are demonstrated on a synthetic and a real life spanner image with various degrees of Gaussian noise. A comparison with the classical PSONN architecture reveals the extraction and time efficiency of the proposed QPSONN architecture.