Journal of Intelligent & Fuzzy Systems - Volume 42, issue 6

Authors: Sun, Guangling | Hu, Haoqi | Zhang, Xinpeng | Lu, Xiaofeng

Article Type: Research Article

Abstract: Universal Adversarial Perturbations(UAPs), which are image-agnostic adversarial perturbations, have been demonstrated to successfully deceive computer vision models. Proposed UAPs in the case of data-dependent, use the internal layers’ activation or the output layer’s decision values as supervision. In this paper, we use both of them to drive the supervised learning of UAP, termed as fully supervised UAP(FS-UAP), and design a progressive optimization strategy to solve the FS-UAP. Specifically, we define an internal layers supervised objective relying on multiple major internal layers’ activation to estimate the deviations of adversarial examples from legitimate examples. We also define an output layer supervised objective …relying on the logits of output layer to evaluate attacking degrees. In addition, we use the UAP found by previous stage as the initial solution of the next stage so as to progressively optimize the UAP stage-wise. We use seven networks and ImageNet dataset to evaluate the proposed FS-UAP, and provide an in-depth analysis for the latent factors affecting the performance of universal attacks. The experimental results show that our FS-UAP (i) has powerful capability of cheating CNNs (ii) has superior transfer-ability across models and weak data-dependent (iii) is appropriate for both untarget and target attacks. Show more

Keywords: Deep learning models, universal adversarial perturbations, fully supervised, progressive optimization

DOI: 10.3233/JIFS-210728

Citation: Journal of Intelligent & Fuzzy Systems, vol. 42, no. 6, pp. 4959-4968, 2022

Authors: ul Haq, Ehtasham | Ahmad, Ishfaq | Hussain, Abid | Almanjahie, Ibrahim M.

Article Type: Research Article

Abstract:  In the present simulation-based study, a novel parent-centric real-coded crossover operator is introduced with a unique probabilistic aspect of the mixture distribution. Moreover, the mixture distribution is a co-integration of double Pareto and Laplace probability distributions with various parameters. The key objective of the newly proposed methodology is to obtained optimal solutions for complex multimodal optimization problems. Hence, for its global comparison, the newly proposed mixture distribution crossover operator (MDX) is compared with double Pareto (DPX), Laplace (LX), and simulated binary (SBX) crossover operators within the conjunction of three mutation operators (MTPM, PM, and NUM). After a descriptive comparison, a …Quade multiple comparison test is also administered to examine its statistical significance. Furthermore, the performance of the genetic algorithm (GA) is also examined on a set of twenty-one unconstraint benchmark functions with diverse features. The empirical results of the simulation-based study reveal that the mixture-based crossover operator obtained a substantial dominance over all considered crossover operators in terms of computational complexity, robustness, scalability, and capability of exploration and exploitation. Moreover, the Quade multiple comparison test also showed a significant superiority with graphical authentication of the performance index (PI). Show more

Keywords: Genetic algorithms, two-component mixture model, real coded crossover, Quade test, Performance Index

DOI: 10.3233/JIFS-210886

Citation: Journal of Intelligent & Fuzzy Systems, vol. 42, no. 6, pp. 4969-4985, 2022

Authors: Piegat, Andrzej | Pluciński, Marcin

Article Type: Research Article

Abstract: The paper presents the inclusion principle of fuzzy arithmetic results. This principle explains what features should have the span of the result of calculations realized with use of the fuzzy arithmetic. If some kind of fuzzy arithmetic provides results that do not comply with this principle, it means that the arithmetic has incomplete reliability, has errors in its theoretical assumptions and should either be verified or rejected. The principle contributes to the ordering of fuzzy arithmetic rules and thus to its practical applicability.

Keywords: Fuzzy arithmetic, interval arithmetic, multidimensional fuzzy arithmetic, principle of results inclusion

DOI: 10.3233/JIFS-210980

Citation: Journal of Intelligent & Fuzzy Systems, vol. 42, no. 6, pp. 4987-4998, 2022

Authors: Gong, Weijun | Wang, Chaoqing | Jia, Jinlu | Qian, Yurong | Fan, Yingying

Article Type: Research Article

Abstract: Facial expression recognition (FER) has been one of the research focuses in recent years due to its significance in human-computer interactions. However, there are still challenges in the field of FER caused by the diversity and variation of facial expressions in real scenes, the singleness of feature type and the lack of enough discriminant features cannot effectively improve the recognition performance. To solve these problems, we propose a Multi-feature Fusion Network (MFNet) with dual-branch based on deep learning. Firstly, the MFNet uses the pyramid parallel multiscale residual network structure with progressive max-pooling of channel attention to extract multi-level facial features …and enhance the discrimination of features; In the meantime, a shallow Gabor convolutional network is designed to enhance the adaptation of learned features to the orientation and scale changes and improve the ability to capture local details features; Finally, the maximum expression features obtained by the above two networks are fused to make more effective expression recognition. Experiments on three public large-scale wild FER datasets (RAF-DB, FERPlus, and AffectNet) show that our MFNet has a superior recognition performance than other recognition methods. Show more

Keywords: Facial expression recognition, multi-feature fusion, feature extraction, deep learning

DOI: 10.3233/JIFS-211021

Citation: Journal of Intelligent & Fuzzy Systems, vol. 42, no. 6, pp. 4999-5011, 2022

Authors: Luo, Minxia | Xu, Donghui

Article Type: Research Article

Abstract:  In this paper, the concept of α  (x , y )-interval-valued pointwise sustaining degree based on the left-continuous t -representable t -norms is put forward. And then, as a general extension based on the interval-valued pointwise sustaining degree, the interval-valued α  (x , y )-full implication triple I method model, the interval-valued α  (x , y )-quintuple implication principle models and the interval-valued α  (x , y )-similarity measure method models are given. Moreover, the interval-valued R -type α  (x , y )-fuzzy reasoning solutions with triple I method, quintuple implication principle and similarity measure method …are given. Some existing results are special cases of the main conclusions in this paper. Show more

Keywords: Interval-valued pointwise staining degree, left-continuous t-representable t-norms, Triple I method, quintuple implication principle, similarity measure method

DOI: 10.3233/JIFS-211076

Citation: Journal of Intelligent & Fuzzy Systems, vol. 42, no. 6, pp. 5013-5029, 2022

Authors: Miao, Yujie | Zhu, Shiping | Huang, Hua | Li, Junxian | Wei, Xiao | Ma, Lingkai | Pu, Jing

Article Type: Research Article

Abstract:  With the development of convolutional neural networks, aiming at the problem of low efficiency and low accuracy in the process of wood species recognition, a recognition method using an improved convolutional neural network is proposed in this article. First, a large-scale wood dataset was constructed based on the WOOD-AUTH dataset and the data collected. Then, a new model named W_IMCNN was constructed based on Inception and mobilenetV3 networks for wood species identification. Experimental results showed that compared with other models, the proposed model had better recognition performance, such as shorter training time and higher recognition accuracy. In the data set …constructed by us, the accuracy of the test set reaches 96.4%. We used WOOD-AUTH dataset to evaluate the model, and the recognition accuracy reached 98.8%. Compared with state-of-the-art methods, the effectiveness of the W_IMCNN were confirmed. Show more

Keywords: Wood species, images, inception, mobileNetV3, convolutional neural networks

DOI: 10.3233/JIFS-211097

Citation: Journal of Intelligent & Fuzzy Systems, vol. 42, no. 6, pp. 5031-5040, 2022

Authors: Tong, Zhao | Chen, Hongjian | Liu, Bilan | Cai, Jinhui | Cai, Shuo

Article Type: Research Article

Abstract: In recent years, solving combinatorial optimization problems involves more complications, high dimensions, and multi-objective considerations. Combining the advantages of other evolutionary algorithms to enhance the performance of a unique evolutionary algorithm and form a new hybrid heuristic algorithm has become a way to strengthen the performance of the algorithm effectively. However, the intelligent hybrid heuristic algorithm destroys the integrity, universality, and robustness of the original algorithm to a certain extent and increases its time complexity. This paper implements a new idea “ML to choose heuristics” (a heuristic algorithm combined with machine learning technology) which uses the Q-learning method to learn …different strategies in genetic algorithm. Moreover, a selection-based hyper-heuristic algorithm is obtained that can guide the algorithm to make decisions at different time nodes to select appropriate strategies. The algorithm is the hybrid strategy using Q-learning on StudGA (HSQ-StudGA). The experimental results show that among the 14 standard test functions, the evolutionary algorithm guided by Q-learning can effectively improve the quality of arithmetic solution. Under the premise of not changing the evolutionary structure of the algorithm, the hyper-heuristic algorithm represents a new method to solve combinatorial optimization problems. Show more

Keywords: Arithmetic solution, combinatorial optimization, genetic algorithm, hyper-heuristic algorithm, Q-learning

DOI: 10.3233/JIFS-211250

Citation: Journal of Intelligent & Fuzzy Systems, vol. 42, no. 6, pp. 5041-5053, 2022

Authors: Xue, Guangming | Lin, Funing | Liu, Heng | Li, Shenggang

Article Type: Research Article

Abstract: This paper explores the prescribed performance tracking control problem of nonlinear systems with triangular structure. To obtain the desired transient performance and precise estimations of uncertain terms, the techniques of neural network control, sliding mode control and composite learning control are incorporated into the proposed control method. The presented control strategy can ensure the tracking error converges to a prescribed small residual set. Compared with the persistent excitation condition required in the conventional adaptive control, the interval excitation condition needed in the proposed control approach is weak, which guarantees that the radial basis function neural networks approximate the unknown nonlinear …terms more accurately. Finally, two simulation examples are exploited to manifest the effectiveness of the proposed approach. Show more

Keywords: Composite learning, prescribed performance, sliding mode control, neural network approximation, prediction error

DOI: 10.3233/JIFS-211310

Citation: Journal of Intelligent & Fuzzy Systems, vol. 42, no. 6, pp. 5055-5067, 2022

Authors: Zheng, Rong | Jia, Heming | Wang, Shuang | Liu, Qingxin

Article Type: Research Article

Abstract:  Slime mould algorithm (SMA) is a new metaheuristic algorithm proposed in 2020, which has attracted extensive attention from scholars. Similar to other optimization algorithms, SMA also has the drawbacks of slow convergence rate and being trapped in local optimum at times. Therefore, the enhanced SMA named as ESMA is presented in this paper for solving global optimization problems. Two effective methods composed of multiple mutation strategy (MMS) and restart mechanism (RM) are embedded into the original SMA. MMS is utilized to increase the population diversity, and the RM is used to avoid the local optimum. To verify the ESMA’s performance, …twenty-three classical benchmark functions are employed, as well as three well-known engineering design problems, including welded beam design, pressure vessel design and speed reducer design. Several famous optimization algorithms are also chosen for comparison. Experimental results show that the ESMA outperforms other optimization algorithms in most of the test functions with faster convergence speed and higher solution accuracy, which indicates the merits of proposed ESMA. The results of Wilcoxon signed-rank test also reveal that ESMA is significant superior to other comparative optimization algorithms. Moreover, the results of three constrained engineering design problems demonstrate that ESMA is better than comparative algorithms. Show more

Keywords: Slime mould algorithm, multiple mutation strategy, restart mechanism, global optimization, optimization algorithm

DOI: 10.3233/JIFS-211408

Citation: Journal of Intelligent & Fuzzy Systems, vol. 42, no. 6, pp. 5069-5083, 2022

Authors: Huang, Junhui | Shao, Dangguo | Liu, Han | Xiang, Yan | Ma, Lei | Yi, Sanli | Xu, Hui

Article Type: Research Article

Abstract: Automatic segmentation of Magnetic Resonance Imaging (MRI), which bases on Residual U-Net (ResU-Net), helps radiologists to quickly assess the condition. However, the ResU-Net structure requires a large number of parameters and storage model space. It is not convenient to apply to mobile MRI device. To solve this problem, Depthwise Separable Convolution and Squeeze-and-Excitation Residual U-Networks (DSRU-Net) is proposed to segment MRI. Squeeze-and-Excitation method is a channel attention mechanism. The proposed method is conducive to simplify ResU-Net model, making ResU-Net more convenient to be applied to mobile MRI device. The fuzzy comprehensive evaluation method, which includes three evaluation factors are that …the required parameters of the model, the value of Dice Similarity Coefficient (DSC), and the value of Hausdorff Distance (HD), is used to evaluate the test results of the proposed method on the MICCAI 2012 Prostate MR Image Segmentation (PROMISE12) challenge dataset and Automatic Cardiac Diagnosis Challenge (ACDC) dataset. The fuzzy comprehensive evaluation values obtained by the proposed method in 5 PROMISE12 samples and 15 ACDC samples are 0.9889 and 0.9652, respectively. Combining the average results of the two datasets, the proposed method has the best effect in balancing the accuracy of segmentation and the amount of model parameters. Show more

Keywords: Depthwise separable convolution, channel attention mechanism, residual U-Net, MRI, segmentation

DOI: 10.3233/JIFS-211424

Citation: Journal of Intelligent & Fuzzy Systems, vol. 42, no. 6, pp. 5085-5095, 2022