ModPSO-CNN: an evolutionary convolution neural network with application to visual recognition
- PDF / 1,497,479 Bytes
- 12 Pages / 595.276 x 790.866 pts Page_size
- 70 Downloads / 270 Views
METHODOLOGIES AND APPLICATION
ModPSO-CNN: an evolutionary convolution neural network with application to visual recognition Shanshan Tu1 · Sadaqat ur Rehman8 Zhongliang Yang5 · Anis Koubaa6,7
· Muhammad Waqas1,3 · Obaid ur Rehman4 · Zubair Shah2 ·
© Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract Training optimization plays a vital role in the development of convolution neural network (CNN). CNNs are hard to train because of the presence of multiple local minima. The optimization problem for a CNN is non-convex, hence, has multiple local minima. If any of the chosen hyper-parameters are not appropriate, it will end up at bad local minima, which leads to poor performance. Hence, proper optimization of the training algorithm for CNN is the key to converge to a good local minimum. Therefore, in this paper, we introduce an evolutionary convolution neural network (ModPSO-CNN) algorithm. The proposed algorithm results in the fusion of modified particle swarm optimization (ModPSO) along with backpropagation (BP) and convolution neural network (CNN). The training of CNN involves ModPSO along with backpropagation (BP) algorithm to encourage performance improvement by avoiding premature convergence and local minima. The ModPSO have adaptive, dynamic and improved parameters, to handle the issues in training CNN. The adaptive and dynamic parameters bring a proper balance between the global and local search ability, while an improved parameter keeps the diversity of the swarm. The proposed ModPSO algorithm is validated on three standard mathematical test functions and compared with three variants of the benchmark PSO algorithm. Furthermore, the performance of the proposed ModPSO-CNN is also compared with other training algorithms focusing on the analysis of computational cost, convergence and accuracy based on a standard problem specific to classification applications, such as CIFAR-10 dataset and face and skin detection dataset. Keywords Particle swarm optimization · Convolution neural network · Backpropagation · Visual recognition
1 Introduction Communicated by V. Loia.
B
Sadaqat ur Rehman [email protected]
The convolution neural network (CNN) obtained benchmark performance in many computer vision applications, such as classification (Rehman et al. 2016, 2018; Bulan et al. 2017), cross-media search (Rehman et al. 2018), recognition (Low et al. 2017; Ding and Tao 2015), and detection (Zhang et al. 2016; Rehman et al. 2017). The key characteristics of CNN deployed in the layers deployed in a feed-forward manner, i.e., the deeper the CNN the efficient and robust will be the results. Moreover, the efficiency of the system further boosts up with the optimization of the training algorithm and a decrease in output classification loss. These characteristics make CNN robust and efficient algorithms for many
1
Faculty of Information Technology, Beijing University of Technology, Beijing, China
2
Division of ICT, College of Science and Engineering, Hamad Bin Khalifa University, Ar-Rayyan, Qatar
3
Fa
Data Loading...
