rss_2.0Journal of Artificial Intelligence and Soft Computing Research FeedSciendo RSS Feed for Journal of Artificial Intelligence and Soft Computing Research of Artificial Intelligence and Soft Computing Research Feed A New Small Object Detection Model for Aerial Images Based on PP-Yoloe<abstract> <title style='display:none'>Abstract</title> <p>Although object detection has achieved great success in the field of computer vision in the past few years, the performance of detecting small objects has not yet achieved ideal results. For instance, UAV aerial photography object detection plays an important role in traffic monitoring and other fields, but it faces some great challenges. The objects in aerial images are mainly small objects, the resolution of whom is low and the feature expression ability of whom is very weak. Information will be lost in high-dimensional feature maps, and this information is very important for the classification and positioning of small objects. The most common way to improve small object detection accuracy is to use high-resolution images, but this incurs additional computational costs. To address the above-mentioned problems, this article proposes a new model SINextNet, which uses a new dilated convolution module SINext block. This module is based on depth-separable convolution, and can improve the receptive field of the model. While extracting small object features, it can combine small object features with background information, greatly improving the feature expression ability of small objects. The experimental results indicate that the method proposed in this paper can achieve advanced performance across multiple aerial datasets.</p> </abstract>ARTICLEtrue Rethinking Lightweight Network for Self-Supervised Monocular Depth Estimation<abstract> <title style='display:none'>Abstract</title> <p>Self-supervised monocular depth estimation has been widely applied in autonomous driving and automated guided vehicles. It offers the advantages of low cost and extended effective distance compared with alternative methods. However, like automated guided vehicles, devices with limited computing resources struggle to leverage state-of-the-art large model structures. In recent years, researchers have acknowledged this issue and endeavored to reduce model size. Model lightweight techniques aim to decrease the number of parameters while maintaining satisfactory performance. In this paper, to enhance the model’s performance in lightweight scenarios, a novel approach to encompassing three key aspects is proposed: (1) utilizing LeakyReLU to involve more neurons in manifold representation; (2) employing large convolution for improved recognition of edges in lightweight models; (3) applying channel grouping and shuffling to maximize the model efficiency. Experimental results demonstrate that our proposed method achieves satisfactory outcomes on KITTI and Make3D benchmarks while having only 1.6M trainable parameters, representing a reduction of 27% compared with the previous smallest model, Lite-Mono-tiny, in monocular depth estimation.</p> </abstract>ARTICLEtrue of Alternatives Described by Atanassov’s Intuitionistic Fuzzy Sets – Reconciling Some Misunderstandings<abstract> <title style='display:none'>Abstract</title> <p>Atanassov’s intuitionistic fuzzy sets (IFSs) are a very convenient tool for describing alternatives/options while making decisions because they make it possible to naturally represent the pros, cons, and hesitation. The IFSs have attracted a significant interest and have been applied in various fields. Of course, their use poses some challenges. One of the main challenges is the ranking of alternatives/options described by the intuitionistic fuzzy sets, to be called for brevity the intuitionistic fuzzy alternatives. This is a crucial issue, notably for the applications, for instance, in decision making. We first present in detail and analyze the benefits of a method we introduced previously (cf. Szmidt and Kacprzyk [<xref ref-type="bibr" rid="j_jaiscr-2024-0013_ref_001">1</xref>]). For this method, we augment the original assumptions with an additional assumption, which is justified and inherently reasonable. As a result, we obtain formulas which are better justified than those previously used as they explicitly consider the arguments in favor (pro), against (con), and hesitance. Since the intuitionistic fuzzy alternatives can not be linearly ranked, then the additional assumptions during the ranking process are necessary. We address these issues and analyze examples to clarify our new approach. We examine some other methods discussed in the literature and analyze their results, and show that the new assumptions reconcile some misconceptions raised by those other papers.</p> </abstract>ARTICLEtrue the Structures of Transformer Neural Networks Using Parallel Simulated Annealing<abstract> <title style='display:none'>Abstract</title> <p>The Transformer is an important addition to the rapidly increasing list of different Artificial Neural Networks (ANNs) suited for extremely complex automation tasks. It has already gained the position of the tool of choice in automatic translation in many business solutions. In this paper, we present an automated approach to optimizing the Transformer structure based upon Simulated Annealing, an algorithm widely recognized for both its simplicity and usability in optimization tasks where the search space may be highly complex. The proposed method allows for the use of parallel computing and time-efficient optimization, thanks to modifying the structure while training the network rather than performing the two one after another. The algorithm presented does not reset the weights after changes in the transformer structure. Instead, it continues the training process to allow the results to be adapted without randomizing all the training parameters. The algorithm has shown a promising performance during experiments compared to traditional training methods without structural modifications. The solution has been released as open-source to facilitate further development and use by the machine learning community.</p> </abstract>ARTICLEtrue Hybrid Equilibrium Optimizer Based on Moth Flame Optimization Algorithm to Solve Global Optimization Problems<abstract> <title style='display:none'>Abstract</title> <p>Equilibrium optimizer (EO) is a novel metaheuristic algorithm that exhibits superior performance in solving global optimization problems, but it may encounter drawbacks such as imbalance between exploration and exploitation capabilities, and tendency to fall into local optimization in tricky multimodal problems. In order to address these problems, this study proposes a novel ensemble algorithm called hybrid moth equilibrium optimizer (HMEO), leveraging both the moth flame optimization (MFO) and EO. The proposed approach first integrates the exploitation potential of EO and then introduces the exploration capability of MFO to help enhance global search, local fine-tuning, and an appropriate balance during the search process. To verify the performance of the proposed hybrid algorithm, the suggested HMEO is applied on 29 test functions of the CEC 2017 benchmark test suite. The test results of the developed method are compared with several well-known metaheuristics, including the basic EO, the basic MFO, and some popular EO and MFO variants. Friedman rank test is employed to measure the performance of the newly proposed algorithm statistically. Moreover, the introduced method has been applied to address the mobile robot path planning (MRPP) problem to investigate its problem-solving ability of real-world problems. The experimental results show that the reported HMEO algorithm is superior to the comparative approaches.</p> </abstract>ARTICLEtrue Incremental Spectral Clustering<abstract><title style='display:none'>Abstract</title> <p>Our previous experiments demonstrated that subsets of collections of (short) documents (with several hundred entries) share a common, normalized in some way, eigenvalue spectrum of combinatorial Laplacian. Based on this insight, we propose a method of incremental spectral clustering. The method consists of the following steps: (1) split the data into manageable subsets, (2) cluster each of the subsets, (3) merge clusters from different subsets based on the eigenvalue spectrum similarity to form clusters of the entire set. This method can be especially useful for clustering methods of complexity strongly increasing with the size of the data sample, like in case of typical spectral clustering. Experiments were performed showing that in fact the clustering and merging of subsets yield clusters close to clustering of the entire dataset. Our approach differs from other research streams in that we rely on the entire set (spectrum) of eigenvalues, whereas the other researchers concentrate on few eigenvectors related to lowest eigenvalues. Such eigenvectors are considered in the literature as of low reliability.</p> </abstract>ARTICLEtrue Dynamic Intention-Aware Recommendation with Attention-Based Context-Aware Item Attributes Modeling<abstract><title style='display:none'>Abstract</title> <p>Recommender systems (RSs) often focus on learning users’ long-term preferences, while the sequential pattern of behavior is ignored. On the other hand, sequential RSs try to predict the next action by exploring relations between items in a user’s last interactions but do not consider the general preference. Recently, the performance of RSs has increased by unifying these two types of paradigms. However, existing methods still have two limitations. First, the user’s behavior uncertainty impedes precise learning of preferences. Second, being unable to understand the semantics of items makes the effect of the same item considered in the same way. These limitations jointly prevent RS from learning multifaceted preferences to capture the actual intentions of users. Existing methods have not properly addressed these problems since they ignore context-aware interactions between the user and item in terms of the links between the user and item attributes and sequential user actions over time. To address these challenges, this paper proposes a novel model, called the Dynamic Intention-Aware Recommendation with attention-based context-aware item attributes modeling (DIARec), which is capable of determining users’ preferences based on their goal intention, taking into account the influence of various item features on user decision-making in their current context. Specifically, to model users’ dynamic intentions, we introduce a dynamic intent-aware module to represent the hierarchical relations between items and their attributes in a given session. Experiments on benchmark datasets indicate that the proposed model DIARec outperforms other state-of-the-art methods.</p> </abstract>ARTICLEtrue Analysis and Anomaly Detection System of Transverse Electron Beam Profile Based on Advanced and Interpretable Deep Learning Architectures<abstract><title style='display:none'>Abstract</title> <p>The National Synchrotron Radiation Center SOLARIS, ranked among the top infrastructures of that type worldwide, is the only one located in Central-Eastern Europe, in Poland. The SOLARIS Center, with six fully operational beamlines, serves as a hub for research across a diverse range of disciplines. This cutting-edge facility fosters innovation in fields like biology, chemistry, and physics as well as material engineering, nanotechnology, medicine, and pharmacology. With its advanced infrastructure and multidisciplinary approach, the SOLARIS Center enables discoveries and pushes the boundaries of knowledge. The most important aspect of such enormous research as well as industry infrastructure is to provide stable working conditions for the users and the conducted projects. Due to its unique properties, problem complexities, and challenges that require advanced approaches, the problem of anomaly detection and automatic analysis of signals for the beam stability assessment is still a huge challenge that has not been fully developed. To increase the effectiveness of centers with advanced research infrastructure we focus on the automatic analysis of transverse beam profiles generated by the Pinhole diagnostic beam-line. Pinhole beamlines are typically installed in the middle and high-energy synchrotrons to thoroughly analyze emitted X-rays and therefore assess electron beam quality. To address the problem we take advantage of AI solutions including up-to-date pre-trained convolutional neural network (CNN) models among others EfficientNetB0-B4-B6, InceptionV3 and DenseNet121. In this research, we propose the benchmark for Pinhole image classification including data preprocessing, model implementation, training process, hyperparameter selection as well as testing phase. Created from scratch database contains over one million transverse beam profile images. The proposed solution, based on the InceptionV3 architecture, classifies pinhole beamline images with 94.1% accuracy and 96.6% precision which is a state-of-the-art result in this research area. Finally, we employed interpretability algorithms to perform an analysis of the models and achieved results.</p> </abstract>ARTICLEtrue Novel Explainable AI Model for Medical Data Analysis<abstract><title style='display:none'>Abstract</title> <p>This research focuses on the development of an explainable artificial intelligence (Explainable AI or XAI) system aimed at the analysis of medical data. Medical imaging and related datasets present inherent complexities due to their high-dimensional nature and the intricate biological patterns they represent. These complexities necessitate sophisticated computational models to decode and interpret, often leading to the employment of deep neural networks. However, while these models have achieved remarkable accuracy, their ”black-box” nature raises legitimate concerns regarding their interpretability and reliability in the clinical context.</p> <p>To address this challenge, we can consider the following approaches: traditional statistical methods, a singular complex neural network, or an ensemble of simpler neural networks. Traditional statistical methods, though transparent, often lack the nuanced sensitivity required for the intricate patterns within medical images. On the other hand, a singular complex neural network, while powerful, can sometimes be too generalized, making specific interpretations challenging. Hence, our proposed strategy employs a hybrid system, combining multiple neural networks with distinct architectures, each tailored to address specific facets of the medical data interpretation challenges.</p> <p>The key components of this proposed technology include a module for anomaly detection within medical images, a module for categorizing detected anomalies into specific medical conditions and a module for generating user-friendly, clinically-relevant interpretations.</p> </abstract>ARTICLEtrue Shuffled Frog Leaping Algorithm with Q-Learning for Distributed Hybrid Flow Shop Scheduling Problem with Energy-Saving<abstract><title style='display:none'>Abstract</title> <p>Energy saving has always been a concern in production scheduling, especially in distributed hybrid flow shop scheduling problems. This study proposes a shuffled frog leaping algorithm with Q-learning (QSFLA) to solve distributed hybrid flow shop scheduling problems with energy-saving(DEHFSP) for minimizing the maximum completion time and total energy consumption simultaneously. The mathematical model is provided, and the lower bounds of two optimization objectives are given and proved. A Q-learning process is embedded in the memeplex search of QSFLA. The state of the population is calculated based on the lower bound. Sixteen search strategy combinations are designed according to the four kinds of global search and four kinds of neighborhood structure. One combination is selected to be used in the memeplex search according to the population state. An energy-saving operator is presented to reduce total energy consumption without increasing the processing time. One hundred forty instances with different scales are tested, and the computational results show that QSFLA is a very competitive algorithm for solving DEHFSP.</p> </abstract>ARTICLEtrue Path Understanding Based on Angle Projections in Field Environments<abstract> <title style='display:none'>Abstract</title> <p>Scene understanding is a core problem for field robots. However, many unsolved problems, like understanding bending paths, severely hinder the implementation due to varying illumination, irregular features and unstructured boundaries in field environments. Traditional three-dimensional(3D) environmental perception from 3D point clouds or fused sensors are costly and account poorly for field unstructured semantic information. In this paper, we propose a new methodology to understand field bending paths and build their 3D reconstruction from a monocular camera without prior training. Bending angle projections are assigned to clusters. Through compositions of their sub-clusters, bending surfaces are estimated by geometric inferences. Bending path scenes are approximated bending structures in the 3D reconstruction. Understanding sloping gradient is helpful for a navigating mobile robot to automatically adjust their speed. Based on geometric constraints from a monocular camera, the approach requires no prior training, and is robust to varying color and illumination. The percentage of incorrectly classified pixels were compared to the ground truth. Experimental results demonstrated that the method can successfully understand bending path scenes, meeting the requirements of robot navigation in an unstructured environment.</p> </abstract>ARTICLEtrue Convolutional Layers in DNN Model Based on Time–Frequency Representation of Emotional Speech<abstract> <title style='display:none'>Abstract</title> <p>The paper describes the relations of speech signal representation in the layers of the convolutional neural network. Using activation maps determined by the Grad-CAM algorithm, energy distribution in the time–frequency space and their relationship with prosodic properties of the considered emotional utterances have been analysed. After preliminary experiments with the expressive speech classification task, we have selected the CQT-96 time–frequency representation. Also, we have used a custom CNN architecture with three convolutional layers in the main experimental phase of the study. Based on the performed analysis, we show the relationship between activation levels and changes in the voiced parts of the fundamental frequency trajectories. As a result, the relationships between the individual activation maps, energy distribution, and fundamental frequency trajectories for six emotional states were described. The results show that the convolutional neural network in the learning process uses similar fragments from time–frequency representation, which are also related to the prosodic properties of emotional speech utterances. We also analysed the relations of the obtained activation maps with time-domain envelopes. It allowed observing the importance of the speech signals energy in classifying individual emotional states. Finally, we compared the energy distribution of the CQT representation in relation to the regions’ energy overlapping with masks of individual emotional states. In the result, we obtained information on the variability of energy distributions in the selected signal representation speech for particular emotions.</p> </abstract>ARTICLEtrue Few-Shot Learning Approach for Covid-19 Diagnosis Using Quasi-Configured Topological Spaces<abstract> <title style='display:none'>Abstract</title> <p>Accurate and efficient COVID-19 diagnosis is crucial in clinical settings. However, the limited availability of labeled data poses a challenge for traditional machine learning algorithms. To address this issue, we propose Turning Point (TP), a few-shot learning (FSL) approach that leverages high-level turning point mappings to build sophisticated representations across previously labeled data. Unlike existing FSL models, TP learns using quasi-configured topological spaces and efficiently combines the outputs of diverse TP learners. We evaluated TPFSL using three COVID-19 datasets and compared it with seven different benchmarks. Results show that TPFSL outperformed the top-performing benchmark models in both one-shot and five-shot tasks, with an average improvement of 4.50% and 4.43%, respectively. Additionally, TPFSL significantly outperformed the ProtoNet benchmark by 12.966% and 11.033% in one-shot and five-shot classification problems across all datasets. Ablation experiments were also conducted to analyze the impact of variables such as TP density, network topology, distance measure, and TP placement. Overall, TPFSL has the potential to improve the accuracy and speed of diagnoses for COVID-19 in clinical settings and can be a valuable tool for medical professionals.</p> </abstract>ARTICLEtrue Operational Neural Networks for The Detection of Atrial Fibrillation<abstract> <title style='display:none'>Abstract</title> <p>Atrial fibrillation is a common cardiac arrhythmia, and its incidence increases with age. Currently, numerous deep learning methods have been proposed for AF detection. However, these methods either have complex structures or poor robustness. Given the evidence from recent studies, it is not surprising to observe the limitations in the learning performance of these approaches. This can be attributed to their strictly homogenous conguration, which solely relies on the linear neuron model. The limitations mentioned above have been addressed by operational neural networks (ONNs). These networks employ a heterogeneous network configuration, incorporating neurons equipped with diverse nonlinear operators. Therefore, in this study, to enhance the detection performance while maintaining computational efficiency, a novel model named multi-scale Self-ONNs (MSSelf-ONNs) was proposed to identify AF. The proposed model possesses a significant advantage and superiority over conventional ONNs due to their self-organization capability. Unlike conventional ONNs, MSSelf -ONNs eliminate the need for prior operator search within the operator set library to find the optimal set of operators. This unique characteristic sets MSSelf -ONNs apart and enhances their overall performance. To validate and evaluate the system, we have implemented the experiments on the well-known MIT-BIH atrial fibrillation database. The proposed model yields total accuracies and kappa coefficients of 98% and 0.95, respectively. The experiment results demonstrate that the proposed model outperform the state-of-the-art deep CNN in terms of both performance and computational complexity.</p> </abstract>ARTICLEtrue for Assessing Generalization of Deep Reinforcement Learning in Parameterized Environments<abstract> <title style='display:none'>Abstract</title> <p>In this work, a study focusing on proposing generalization metrics for Deep Reinforcement Learning (DRL) algorithms was performed. The experiments were conducted in DeepMind Control (DMC) benchmark suite with parameterized environments. The performance of three DRL algorithms in selected ten tasks from the DMC suite has been analysed with existing generalization gap formalism and the proposed ratio and decibel metrics. The results were presented with the proposed methods: average transfer metric and plot for environment normal distribution. These efforts allowed to highlight major changes in the model’s performance and add more insights about making decisions regarding models’ requirements.</p> </abstract>ARTICLEtrue Ensuring Software Interoperability Between Deep Learning Frameworks<abstract><title style='display:none'>Abstract</title> <p>With the widespread of systems incorporating multiple deep learning models, ensuring interoperability between target models has become essential. However, due to the unreliable performance of existing model conversion solutions, it is still challenging to ensure interoperability between the models developed on different deep learning frameworks. In this paper, we propose a systematic method for verifying interoperability between pre- and post-conversion deep learning models based on the validation and verification approach. Our proposed method ensures interoperability by conducting a series of systematic verifications from multiple perspectives. The case study confirmed that our method successfully discovered the interoperability issues that have been reported in deep learning model conversions.</p> </abstract>ARTICLEtrue Attack Detection Method for Imbalanced Data in Industrial Cyber-Physical Systems<abstract><title style='display:none'>Abstract</title> <p>Integrating industrial cyber-physical systems (ICPSs) with modern information technologies (5G, artificial intelligence, and big data analytics) has led to the development of industrial intelligence. Still, it has increased the vulnerability of such systems regarding cybersecurity. Traditional network intrusion detection methods for ICPSs are limited in identifying minority attack categories and suffer from high time complexity. To address these issues, this paper proposes a network intrusion detection scheme, which includes an information-theoretic hybrid feature selection method to reduce data dimensionality and the ALLKNN-LightGBM intrusion detection framework. Experimental results on three industrial datasets demonstrate that the proposed method outperforms four mainstream machine learning methods and other advanced intrusion detection techniques regarding accuracy, F-score, and run time complexity.</p> </abstract>ARTICLEtrue Explainable AI Approach to Agrotechnical Monitoring and Crop Diseases Prediction in Dnipro Region of Ukraine<abstract><title style='display:none'>Abstract</title> <p>The proliferation of computer-oriented and information digitalisation technologies has become a hallmark across various sectors in today’s rapidly evolving environment. Among these, agriculture emerges as a pivotal sector in need of seamless incorporation of high-performance information technologies to address the pressing needs of national economies worldwide. The aim of the present article is to substantiate scientific and applied approaches to improving the efficiency of computer-oriented agrotechnical monitoring systems by developing an intelligent software component for predicting the probability of occurrence of corn diseases during the full cycle of its cultivation. The object of research is non-stationary processes of intelligent transformation and predictive analytics of soil and climatic data, which are factors of the occurrence and development of diseases in corn. The subject of the research is methods and explainable AI models of intelligent predictive analysis of measurement data on the soil and climatic condition of agricultural enterprises specialised in growing corn. The main scientific and practical effect of the research results is the development of IoT technologies for agrotechnical monitoring through the development of a computer-oriented model based on the ANFIS technique and the synthesis of structural and algorithmic provision for identifying and predicting the probability of occurrence of corn diseases during the full cycle of its cultivation.</p> </abstract>ARTICLEtrue New Approach to Detecting and Preventing Populations Stagnation Through Dynamic Changes in Multi-Population-Based Algorithms<abstract><title style='display:none'>Abstract</title> <p>In this paper, a new mechanism for detecting population stagnation based on the analysis of the local improvement of the evaluation function and the infinite impulse response filter is proposed. The purpose of this mechanism is to improve the population stagnation detection capability for various optimization scenarios, and thus to improve multi-population-based algorithms (MPBAs) performance. In addition, various other approaches have been proposed to eliminate stagnation, including approaches aimed at both improving performance and reducing the complexity of the algorithms. The developed methods were tested, among the others, for various migration topologies and various MPBAs, including the MNIA algorithm, which allows the use of many different base algorithms and thus eliminates the need to select the population-based algorithm for a given simulation problem. The simulations were performed for typical benchmark functions and control problems. The obtained results confirm the validity of the developed method.</p> </abstract>ARTICLEtrue Uncertainty Quantification For Offline Reinforcement Learning<abstract><title style='display:none'>Abstract</title> <p>In many Reinforcement Learning (RL) tasks, the classical online interaction of the learning agent with the environment is impractical, either because such interaction is expensive or dangerous. In these cases, previous gathered data can be used, arising what is typically called Offline RL. However, this type of learning faces a large number of challenges, mostly derived from the fact that exploration/exploitation trade-off is overshadowed. In addition, the historical data is usually biased by the way it was obtained, typically, a sub-optimal controller, producing a distributional shift from historical data and the one required to learn the optimal policy. In this paper, we present a novel approach to deal with the uncertainty risen by the absence or sparse presence of some state-action pairs in the learning data. Our approach is based on shaping the reward perceived from the environment to ensure the task is solved. We present the approach and show that combining it with classic online RL methods make them perform as good as state of the art Offline RL algorithms such as CQL and BCQ. Finally, we show that using our method on top of established offline learning algorithms can improve them.</p> </abstract>ARTICLEtrue