Designing pulse-coupled neural networks with spike-synchronization-dependent plasticity rule: image segmentation and mem
- PDF / 1,736,030 Bytes
- 12 Pages / 595.276 x 790.866 pts Page_size
- 103 Downloads / 247 Views
(0123456789().,-volV)(0123456789(). ,- volV)
ORIGINAL ARTICLE
Designing pulse-coupled neural networks with spike-synchronizationdependent plasticity rule: image segmentation and memristor circuit application Xudong Xie1 • Shiping Wen2 • Zheng Yan3 • Tingwen Huang4 • Yiran Chen5 Received: 21 May 2019 / Accepted: 23 January 2020 Ó Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract Pulse-coupled neural network (PCNN) is a powerful unsupervised learning model with many parameters to be determined empirically. In particular, the weight matrix is invariable in the iterative process, which is inconsistent with the actual biological system. Based on the existing research foundation of biology and neural network, we propose a spike-synchronization-dependent plasticity (SSDP) rule. In this paper, the mathematical model and algorithm of SSDP are presented. Furthermore, a novel memristor-based circuit model of SSDP is designed. Finally, experimental results demonstrate that SSDP has greatly improved the image processing capabilities of PCNN. Keywords Spike-synchronization-dependent plasticity Pulse-coupled neural network Memristor-based circuit Biomedical image segmentation
1 Introduction Pulse-coupled neural network simulates the behavior of visual cortical optic nerve cells in mammals such as cats [27, 28, 69]. PCNN utilizes the unique linear additive and nonlinear multiplicative modulation coupling characteristics of neurons, considering the delay characteristics and exponential decay characteristics of bioelectric transmission. Other than that, it describes the characteristics of the oscillation pulse sequence generated by the simultaneous
& Shiping Wen [email protected] 1
School of Electronic Information and Communications, Huazhong University of Science and Technology, Wuhan 430074, China
2
Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu, China
3
Centre for Artificial Intelligence, University of Technology Sydney, Ultimo 2007, Australia
4
Science Program, Texas A & M University at Qatar, 23874 Doha, Qatar
5
Department of Electrical and Computer Engineering, Duke University, Durham, NC 27708, USA
excitation of adjacent connected neurons when the visual field of the mammalian optic nerve system is properly stimulated. The bias in the internal state of a PCNN neuron is actually an equivalent representation of the internal equilibrium state when the neuron is in a suppressed state. PCNN is a single-layer neural network with two-dimensional matrix structure, whose size and layout depends on the input image. It can perform image fusion [25, 50], image segmentation [17, 46] and object detection [8, 36] without training process, so it is promising for real-time image processing. To enable a PCNN model work properly, many hyperparameters need to be determined. However, the relationship between the model hyper-parameters and image processing effect is not clear. Therefore, the adjustment of model parameters lacks theoretical basis, and i
Data Loading...
