Article
Computer Science, Artificial Intelligence
Dheeb Albashish, Abdelaziz Hammouri, Malik Braik, Jaffar Atwan, Shahnorbanun Sahran
Summary: The study proposed a hybrid metaheuristic model based on BBO-SVM-RFE for feature selection, which outperformed other methods in terms of accuracy and number of selected features. Results revealed the high potential of BBO-SVM-RFE in reliably searching the feature space to obtain the optimal combination of features.
APPLIED SOFT COMPUTING
(2021)
Article
Orthopedics
Hidetoshi Nakao, Masakazu Imaoka, Mitsumasa Hida, Ryota Imai, Misa Nakamura, Kazuyuki Matsumoto, Kenji Kita
Summary: This study used support vector machine-recursive feature elimination (SVM-RFE) to identify the factors related to hallux valgus (HV) and their importance. Analysis of data from 864 participants aged ≥ 18 years showed that age, sex, and body weight were associated with HV.
BMC MUSCULOSKELETAL DISORDERS
(2023)
Article
Chemistry, Multidisciplinary
Haoxiang Xu, Tongyao Ren, Zhuangda Mo, Xiaohui Yang
Summary: This research proposes a fault diagnostic approach for the chemical industry using a probabilistic neural network based on feature selection and a bio-heuristic optimizer. The findings demonstrate the advantages of the proposed method in simplifying and eliminating redundant features, as well as optimization capabilities.
APPLIED SCIENCES-BASEL
(2022)
Article
Plant Sciences
Aida Shomali, Sasan Aliniaeifard, Mohammad Reza Bakhtiarizadeh, Mahmoud Lotfi, Mohammad Mohammadian, Mohammad Sadegh Vafaei Sadi, Anshu Rastogi
Summary: High light stress directly affects the photosynthesis apparatus, making breeding plants with tolerance against this stress highly demanded. Chlorophyll fluorescence can be used to indicate plant stress and was compared in plants exposed to high light and control conditions. Artificial neural network algorithms were applied to identify reliable features for screening plant tolerance against high light. The selected features were then used to categorize tomato genotypes and validated using measurements of foliar hydrogen peroxide and malondialdehyde contents.
PLANT PHYSIOLOGY AND BIOCHEMISTRY
(2023)
Article
Biology
Asad Khan, Hafeez Ur Rehman, Usman Habib, Umer Ijaz
Summary: N6-methyladenosine (m6A) is a common post-transcription modification in cellular RNA, playing important roles in regulating various biological processes and being associated with different diseases. However, accurate detection of m6A sites is challenging for existing predictors, especially in complex patterns. To address this, a novel predictor called m6A-Finder is proposed, which creates features based on both global and local sequence order and utilizes the mRMR algorithm to remove redundant features and prevent overfitting. Evaluations on the widely used Saccharomyces cerevisiae dataset demonstrate that m6A-Finder achieves high accuracy, sensitivity, specificity, and a significant Matthew correlation coefficient value.
COMPUTATIONAL BIOLOGY AND CHEMISTRY
(2022)
Article
Mathematics
Md. Humaun Kabir, Shabbir Mahmood, Abdullah Al Shiam, Abu Saleh Musa Miah, Jungpil Shin, Md. Khademul Islam Molla
Summary: Analyzing EEG signals with machine learning for BCI has gained attention. Researchers have been developing MI-based BCI systems but face challenges due to irrelevant features and high complexity. Feature selection is crucial to overcome these issues.
Article
Multidisciplinary Sciences
Abdullah Al-Saleh
Summary: The Internet of Things field has posed challenges for network architectures, with cybersecurity being the primary goal for intrusion detection systems (IDSs). To improve IDS performance, researchers have focused on efficiently protecting connected data and devices due to the increasing number and types of attacks. This paper presents a novel IDS model that offers accurate detection in less processing time by reducing computational complexity. The model uses the Gini index method to determine security feature impurity and improve selection processes. It also employs a balanced communication-avoiding support vector machine decision tree method for enhanced intrusion detection accuracy. Evaluation using the UNSW-NB 15 dataset shows that the proposed model achieves a high attack detection performance of approximately 98.5%.
SCIENTIFIC REPORTS
(2023)
Article
Environmental Sciences
Muhammed Yildirim, Ahmet cinar, Emine CengIl
Summary: In this study, a hybrid model was developed for weather image classification using artificial intelligence methods. By combining and optimizing features extracted by CNN models and classifying them with SVM, the hybrid model outperformed existing pre-trained architectures, demonstrating the success of feature concatenation using CNN architectures for classification.
GEOCARTO INTERNATIONAL
(2022)
Article
Engineering, Electrical & Electronic
Vikas Singh, Nishchal K. Verma
Summary: In recent years, intelligent condition-based monitoring of rotary machinery systems has become a major research focus. Using mRMR and deep learning models can improve fault diagnostics performance by reducing data redundancy and decreasing data dependency for training the model. The proposed frameworks show better diagnostic accuracy and faster processing of data with many features.
IEEE SENSORS JOURNAL
(2021)
Article
Spectroscopy
Xiang Song-yang, Xu Zhang-hua, Zhang Yi-wei, Zhang Qi, Zhou Xin, Yu Hui, Li Bin, Li Yi-fan
Summary: This paper proposes a ReliefF-RFE feature selection algorithm for hyperspectral image classification, which combines the filtered ReliefF algorithm and the wrapped recursive feature elimination algorithm (RFE). Experimental results show that the ReliefF-RFE algorithm can reduce feature dimension and operation time while ensuring classification accuracy.
SPECTROSCOPY AND SPECTRAL ANALYSIS
(2022)
Article
Automation & Control Systems
Muhammad Arif, Saeed Ahmed, Fang Ge, Muhammad Kabir, Yaser Daanial Khan, Dong-Jun Yu, Maha Thafar
Summary: A novel predictor called Stack-ACPred has been developed for the accurate identification of anticancer peptides (ACPs). This method combines three feature encoding strategies and utilizes an optimization algorithm for feature fusion and attribute selection, resulting in the construction of an effective ensemble model. Empirical results demonstrate the excellent discriminative power of this method for annotating large scale ACPs.
CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS
(2022)
Article
Chemistry, Analytical
Kiril Alexiev, Teodor Vakarelski
Summary: This paper investigates microsaccades as a source of information for biometric analysis and explains why they are preferred over other fixational eye movements. The process of microsaccade extraction is described, including the definition and derivation of thirteen parameters for analysis. A gradient algorithm is used to solve the biometric problem and an assessment of parameter weights is made.
Article
Computer Science, Software Engineering
Duygu Kaya
Summary: Gender and Parkinson disease can be simultaneously detected using the proposed method, which achieves a high accuracy through feature extraction and selection, providing valuable insights for related research.
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE
(2022)
Article
Genetics & Heredity
Ling-Fang Ye, Jia-Yi Weng, Li-Da Wu
Summary: This study aimed to investigate the molecular subgroups of dilated cardiomyopathy (DCM). Based on gene expression profiles, heart tissue samples from DCM patients were clustered into three molecular subgroups. Each subgroup had its specific gene modules that were closely associated with the occurrence and progression of DCM. Furthermore, different molecular subgroups exhibited distinct clinical characteristics, highlighting the need for personalized treatment in DCM patients.
FRONTIERS IN GENETICS
(2023)
Article
Chemistry, Multidisciplinary
Chengzhe Lv, Yuefeng Lu, Miao Lu, Xinyi Feng, Huadan Fan, Changqing Xu, Lei Xu
Summary: In this study, a feature dimension reduction algorithm combining Fisher Score and mRMR feature selection method was proposed and tested on GF-2 remote sensing imagery. The experimental analysis showed that this method provides higher accuracy in remote sensing image classification.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Interdisciplinary Applications
Boyan Zhang, Zhiyong Wang, Junbin Gao, Chantal Rutjes, Kaitlin Nufer, Dacheng Tao, David Dagan Feng, Scott W. Menzies
Summary: Short-term monitoring of lesion changes in melanoma screening is currently heavily dependent on individual clinicians' experience and bias, leading to subjective decisions. This paper introduces a novel deep learning-based method for automatically detecting short-term lesion changes, using a Siamese structure and Tensorial Regression Process to improve accuracy. Experimental results on a large dataset show promising results for objective melanoma screening.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2021)
Article
Computer Science, Hardware & Architecture
Ruibin Gu, Qiuxia Wu, Yuqiong Li, Wenxiong Kang, Wing W. Y. Ng, Zhiyong Wang
Summary: In this paper, a novel rotation-invariant network named ELGANet is proposed to tackle the issues of rotation disturbance and insufficient labeled data. The ELGANet includes enhanced local representation learning module and global alignment module to capture geometric relationship and adaptively generate rotation-invariant coordinates. Besides, an unsupervised learning network ELGANet-U is also introduced to generate discriminative and rotation-invariant representation without human supervision.
Article
Computer Science, Artificial Intelligence
Mingyang Ma, Shaohui Mei, Shuai Wan, Zhiyong Wang, Xian-Sheng Hua, David Dagan Feng
Summary: In this paper, a general framework called graph convolutional dictionary selection with L-2, L-p (0 < p <= 1) norm (GCDS(2,p)) is proposed for both keyframe selection and skimming based summarization in video summarization. The structured information in videos is taken into account by incorporating graph embedding into dictionary selection. L-2, L-p (0 < p <= 1) norm constrained row sparsity with flexible p values is used for selecting diverse and representative keyframes or key shots.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2022)
Article
Engineering, Electrical & Electronic
Shan Cao, Gaoyun An, Zhenxing Zheng, Zhiyong Wang
Summary: In this paper, a Vision-enhanced and Consensus-aware Transformer (VCT) is proposed for image captioning. The model extends the self-attention module and introduces memory-based attention and visual perception modules to enhance visual representation of images. Consensus knowledge is learned through word correlation graph and graph convolutional network. Experimental results demonstrate state-of-the-art performance on two benchmark datasets.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2022)
Review
Agronomy
Guy R. Y. Coleman, Asher Bender, Kun Hu, Shaun M. Sharpe, Arnold W. Schumann, Zhiyong Wang, Muthukumar V. Bagavathiannan, Nathan S. Boyd, Michael J. Walsh
Summary: Advances in weed recognition technologies over the past 50 years have provided the necessary performance for site-specific weed control in large-scale production systems. These technologies offer improved management of diverse weed morphology and enable the use of nonselective weed control options such as lasers and electrical weeding. Recent research has focused on computer vision techniques and deep convolutional neural network (CNN) approaches for weed recognition.
Article
Computer Science, Interdisciplinary Applications
Kun Hu, Wenhua Wu, Wei Li, Milena Simic, Albert Zomaya, Zhiyong Wang
Summary: A novel deep learning architecture, A-ENN, is proposed for longitudinal grading of knee osteoarthritis (KOA) severity. By obtaining evolution traces through an adversarial training scheme, the fine-grained domain knowledge is fused with general convolutional image representations, achieving longitudinal grading.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2022)
Article
Computer Science, Artificial Intelligence
Shan Zeng, Xiangjun Duan, Hao Li, Jun Bai, Yuanyan Tang, Zhiyong Wang
Summary: In this article, a novel robust and sparse possibilistic K-subspace (RSPKS) clustering algorithm is proposed to handle clustering of noisy, high-dimensional, and structurally complex data. The algorithm integrates subspace recovery and possibilistic clustering algorithms under a unified sparse framework to effectively deal with the adverse impact of noisy samples and complex data structures. Experimental results on both synthetic and real-world datasets demonstrate that the proposed method outperforms state-of-the-art algorithms in terms of clustering accuracy.
IEEE TRANSACTIONS ON FUZZY SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Kun Hu, Zhiyong Wang, Kaylena A. Ehgoetz Martens, Markus Hagenbuchner, Mohammed Bennamoun, Ah Chung Tsoi, Simon J. G. Lewis
Summary: This study proposes a multimodal learning-based FoG detection method using a graph fusion neural network (GFN) that combines footstep pressure maps and video recordings. The GFN constructs multimodal graphs to reduce redundancy among different modalities and achieves superior performance. Experimental results show promising FoG detection with an AUC of 0.882.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Interdisciplinary Applications
Wenxi Yue, Hongen Liao, Yong Xia, Vincent Lam, Jiebo Luo, Zhiyong Wang
Summary: This paper proposes a Cascade Multi-Level Transformer Network (CMTNet) for recognizing surgical phases, and introduces the Adaptive Multi-Level Context Aggregation (AMCA) modules. Through the gradual enrichment of multi-level semantics and the refinement of key context, CMTNet achieves more accurate phase prediction.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2023)
Article
Computer Science, Information Systems
Kun Hu, Shaohui Mei, Wei Wang, Kaylena A. Ehgoetz Martens, Liang Wang, Simon J. G. Lewis, David D. Feng, Zhiyong Wang
Summary: Freezing of gait (FoG) is a common symptom of Parkinson's disease, and a computer-aided detection and quantification tool for FoG is important for improving treatment quality. Footstep pressure sequences obtained from pressure sensitive gait mats provide a non-invasive way to evaluate FoG, and the proposed Adversarial Spatio-temporal Network (ASTN) is a novel deep learning architecture that can learn FoG patterns and achieve robust detection. In experiments, ASTN outperformed conventional learning methods with an AUC of 0.85.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Puning Yang, Huaibo Huang, Zhiyong Wang, Aijing Yu, Ran He
Summary: In this paper, a novel contrastive distillation calibration (CDC) framework is proposed to address the issue of model generalization in face forgery detection. The framework distills contrastive representations with confidence calibration. A dual-teacher module is devised to separately learn knowledge for each forgery type, and a contrastive representation learning strategy is presented to enhance diverse forgery artifacts. Moreover, label smoothing is introduced to calibrate the model confidence with the target outputs.
COMPUTER VISION - ACCV 2022, PT IV
(2023)
Article
Chemistry, Physical
Peng-Fei Yang, Wei Shu
Summary: Stereogenic carbon centers with C(sp3)-C(sp3) bonds are widely present in natural products, bioactive targets, and chiral organic materials. Transition-metal-catalyzed C(sp3)-C(sp3) bond-forming processes offer a promising solution to generate such stereogenic centers. Recent progress in the in situ formation of alkyl metallic reagents enabled by hydrometallation of olefins for asymmetric alkyl-alkyl cross-coupling is highlighted. Mechanistic considerations, challenges, and future efforts in asymmetric hydroalkylation of olefins are also discussed.
Article
Computer Science, Artificial Intelligence
Renfei Sun, Kun Hu, Kaylena A. Ehgoetz Martens, Markus Hagenbuchner, Ah Chung Tsoi, Mohammed Bennamoun, Simon J. G. Lewis, Zhiyong Wang
Summary: Freezing of Gait (FoG) is a common symptom of Parkinson's disease and machine learning-based methods can effectively detect it. This article proposes a novel deep learning architecture called higher order polynomial transformer (HP-Transformer) for fine-grained FoG detection based on vision inputs. The proposed method incorporates pose and appearance feature sequences and achieves an AUC of 0.92 for FoG detection.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Peiqin Zhuang, Yu Guo, Zhipeng Yu, Luping Zhou, Lei Bai, Ding Liang, Zhiyong Wang, Yali Wang, Wanli Ouyang
Summary: Motion modeling plays a crucial role in modern action recognition methods. However, variations in motion dynamics across different video clips present a challenge in adaptively covering proper motion information. In this paper, we propose a Motion Diversification and Selection (MoDS) module that generates diversified spatio-temporal motion features and dynamically selects the suitable motion representation for categorizing input videos. Our method achieves state-of-the-art performance on benchmarks with large motion variations.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2022)
Article
Computer Science, Artificial Intelligence
Jiahao Xu, Boyan Zhang, Zhiyong Wang, Yang Wang, Fang Chen, Junbin Gao, David Dagan Feng
Summary: Public speaking is a crucial skill in daily communication. The lack of personalized feedback hinders the improvement of this skill, even with more practice. This research proposes a novel convolutional clustering neural network (CCNN) to solve the problem of personalized feedback by learning from online public speech videos. Experimental results on a self-built affective audio annotation dataset show that our proposed method outperforms traditional CNN-based approaches, achieving better affective annotation with a lower hamming loss.
IEEE TRANSACTIONS ON AFFECTIVE COMPUTING
(2022)
Article
Computer Science, Artificial Intelligence
Guang-Hai Liu, Zuo-Yong Li, Jing-Yu Yang, David Zhang
Summary: This article introduces a novel image retrieval method that improves retrieval performance by using sublimated deep features. The method incorporates orientation-selective features and color perceptual features, effectively mimicking these mechanisms to provide a more discriminating representation.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Fengguang Peng, Zihan Ding, Ziming Chen, Gang Wang, Tianrui Hui, Si Liu, Hang Shi
Summary: RGB-Thermal (RGB-T) semantic segmentation is an emerging task that aims to improve the robustness of segmentation methods under extreme imaging conditions by using thermal infrared modality. The challenges of foreground-background distinguishment and complementary information mining are addressed by proposing a cross modulation process with two collaborative components. Experimental results show that the proposed method achieves state-of-the-art performances on current RGB-T segmentation benchmarks.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Baihong Han, Xiaoyan Jiang, Zhijun Fang, Hamido Fujita, Yongbin Gao
Summary: This paper proposes a novel automatic prompt generation method called F-SCP, which focuses on generating accurate prompts for low-accuracy classes and similar classes. Experimental results show that our approach outperforms state-of-the-art methods on six multi-domain datasets.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Huikai Liu, Ao Zhang, Wenqian Zhu, Bin Fu, Bingjian Ding, Shengwu Xiong
Summary: Adverse weather conditions present challenges for computer vision tasks, and image de-weathering is an important component of image restoration. This paper proposes a multi-patch skip-forward structure and a Residual Deformable Convolutional module to improve feature extraction and pixel-wise reconstruction.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Oliver M. Crook, Mihai Cucuringu, Tim Hurst, Carola-Bibiane Schonlieb, Matthew Thorpe, Konstantinos C. Zygalakis
Summary: The transportation LP distance (TLP) is a generalization of the Wasserstein WP distance that can be applied directly to color or multi-channelled images, as well as multivariate time-series. TLP interprets signals as functions, while WP interprets signals as measures. Although both distances are powerful tools in modeling data with spatial or temporal perturbations, their computational cost can be prohibitively high for moderate pattern recognition tasks. The linear Wasserstein distance offers a method for projecting signals into a Euclidean space, and in this study, we propose linear versions of the TLP distance (LTLP) that show significant improvement over the linear WP distance in signal processing tasks while being several orders of magnitude faster to compute than the TLP distance.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Haitao Tian, Shiru Qu, Pierre Payeur
Summary: This paper proposes a method of target-dependent classifier, which optimizes the joint hypothesis of domain adaptation into a target-dependent hypothesis that better fits with the target domain clusters through an unsupervised fine-tuning strategy and the concept of meta-learning. Experimental results demonstrate that this method outperforms existing techniques in synthetic-to-real adaptation and cross-city adaptation benchmarks.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Qingsen Yan, Axi Niu, Chaoqun Wang, Wei Dong, Marcin Wozniak, Yanning Zhang
Summary: Deep learning-based methods have achieved remarkable results in the field of super-resolution. However, the limitation of paired training image sets has led researchers to explore self-supervised learning. However, the assumption of inaccurate downscaling kernel functions often leads to degraded results. To address this issue, this paper introduces KGSR, a kernel-guided network that trains both upscaling and downscaling networks to generate high-quality high-resolution images even without knowing the actual downscaling process.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Yifan Chen, Xuelong Li
Summary: Gait recognition is a popular technology for identification due to its ability to capture gait features over long distances without cooperation. However, current methods face challenges as they use a single network to extract both temporal and spatial features. To solve this problem, we propose a two-branch network that focuses on spatial and temporal feature extraction separately. By combining these features, we can effectively learn the spatio-temporal information of gait sequences.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Wei Shi, Wentao Zhang, Wei-shi Zheng, Ruixuan Wang
Summary: This article proposes a simple yet effective visualization framework called PAMI, which does not require detailed model structure and parameters to obtain visualization results. It can be applied to various prediction tasks with different model backbones and input formats.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Xiaobo Hu, Jianbo Su, Jun Zhang
Summary: This paper reviews the latest technologies in pattern recognition, highlighting their instabilities and failures in practical applications. From a control perspective, the significance of disturbance rejection in pattern recognition is discussed, and the existing problems are summarized. Finally, potential solutions related to the application of compensation on features are discussed to emphasize future research directions.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Andres Felipe Posada-Moreno, Nikita Surya, Sebastian Trimpe
Summary: Convolutional neural networks are widely used in critical systems, and explainable artificial intelligence has proposed methods for generating high-level explanations. However, these methods lack the ability to determine the location of concepts. To address this, we propose a novel method for automatic concept extraction and localization based on pixel-wise aggregations, and validate it using synthetic datasets.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Peng Bao, Jianian Li, Rong Yan, Zhongyi Liu
Summary: In this paper, a novel Dynamic Graph Contrastive Learning framework, DyGCL, is proposed to capture the temporal consistency in dynamic graphs and achieve good performance in node representation learning.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Kristian Schultz, Saptarshi Bej, Waldemar Hahn, Markus Wolfien, Prashant Srivastava, Olaf Wolkenhauer
Summary: Research indicates that deep generative models perform poorly compared to linear interpolation-based methods for synthetic data generation on small, imbalanced tabular datasets. To address this, a new approach called ConvGeN, combining convex space learning with deep generative models, has been proposed. ConvGeN improves imbalanced classification on small datasets while remaining competitive with existing linear interpolation methods.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Khondaker Tasrif Noor, Antonio Robles-Kelly
Summary: In this paper, the authors propose H-CapsNet, a capsule network designed for hierarchical image classification. The network effectively captures hierarchical relationships using dedicated capsules for each class hierarchy. A modified hinge loss is utilized to enforce consistency among the involved hierarchies. Additionally, a strategy for dynamically adjusting training parameters is presented to achieve better balance between the class hierarchies. Experimental results demonstrate that H-CapsNet outperforms competing hierarchical classification networks.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Lei Liu, Guorun Li, Yuefeng Du, Xiaoyu Li, Xiuheng Wu, Zhi Qiao, Tianyi Wang
Summary: This study proposes a new agricultural image segmentation model called CS-Net, which uses Simple-Attention Block and Simpleformer to improve accuracy and inference speed, and addresses the issue of performance collapse of Transformers in agricultural image processing.
PATTERN RECOGNITION
(2024)