Article
Computer Science, Information Systems
Muhammad Aminu, Noor Atinah Ahmad
Summary: By incorporating a locality preserving feature, LPPLSDA enhances the performance of partial least squares discriminant analysis, especially in face recognition tasks. Experimental results consistently show that LPPLSDA outperforms the conventional PLS-DA method on various benchmarked face databases.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Ruisheng Ran, Yinshan Ren, Shougui Zhang, Bin Fang
Summary: In this paper, a novel dimensionality reduction method named PDLPP is proposed, which addresses the small-sample-size problem of the DLPP method and achieves better pattern classification performance through nonlinear mapping. Experimental results demonstrate the superiority of PDLPP over state-of-the-art methods.
JOURNAL OF MATHEMATICAL IMAGING AND VISION
(2021)
Article
Engineering, Electrical & Electronic
Yan-Lin He, Kun Li, Ning Zhang, Yuan Xu, Qun-Xiong Zhu
Summary: The article proposes a fault diagnosis methodology using discrimination locality preserving projections integrated with sparse autoencoder (SAEDLPP) for achieving higher accuracy in fault diagnosis in industrial processes.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2021)
Article
Chemistry, Analytical
Dingzhong Feng, Shanyu He, Zihao Zhou, Ye Zhang
Summary: This paper proposes a novel feature extraction method called principal component local preservation projections (PCLPP) for finger vein recognition. The method combines principal component analysis (PCA) and locality preserving projections (LPP) to construct a projection matrix that preserves both global and local features of the image, thereby improving the accuracy of image recognition.
Article
Computer Science, Artificial Intelligence
Tingting Su, Dazheng Feng, Haoshuang Hu, Meng Wang, Mohan Chen
Summary: This study proposes a locality-preserving triplet discriminative projections algorithm to address the challenges of neighbor selection and intrinsic structure disruption in graph embedding. The algorithm constructs locality-preserving and discriminative graphs to enhance separability and preserve local structures. Experimental results demonstrate its superiority over other dimensionality reduction methods.
Article
Computer Science, Artificial Intelligence
Jianyong Zhu, Jingwei Chen, Bin Xu, Hui Yang, Feiping Nie
Summary: This paper proposes a method called OLPPFS, which preserves the local geometric structure within the feature subspace by imposing the 2,0-norm pound constraint. The graph-embedding learning method is used to accelerate the construction of a sparsity affinity graph. Experimental results demonstrate that the proposed method is superior to others in preserving the local geometric structure of the dataset with less time consumption.
Article
Automation & Control Systems
Ning Zhang, Yuan Xu, Qun-Xiong Zhu, Yan-Lin He
Summary: This article presents a novel dimensionality reduction algorithm named DPNLP for fault diagnosis. To solve the singular matrix problem, a regularization-based version of DPNLP called RDPNLP is also introduced. Simulation results show that RDPNLP outperforms other related methods in fault diagnosis.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Engineering, Electrical & Electronic
Qun-Xiong Zhu, Xin-Wei Wang, Ning Zhang, Yuan Xu, Yan-Lin He
Summary: A novel K-medoids-based synthetic minority oversampling technique (KMS-LPP) is proposed for fault diagnosis in industrial processes. The method generates minority fault samples and reduces the dimensionality of data to enhance fault diagnosis performance.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Article
Mathematics
Minghua Wan, Yuxi Zhang, Guowei Yang, Hongjian Guo
Summary: The 2DESDLPP algorithm addresses the small sample size problem and reduces redundant information by combining matrix exponential function and elastic net regression with the 2DDLPP algorithm. It effectively preserves feature information and demonstrates higher accuracy rates compared to other mainstream feature extraction algorithms.
Article
Computer Science, Hardware & Architecture
Ning Zhang, Yuan Xu, Qun-Xiong Zhu, Yan-Lin He
Summary: This article proposes an improved locality preserving projections method based on the heat-kernel and cosine weight matrix, named HC-LPP, for fault diagnosis. By optimizing the weight matrix, HC-LPP considers both the distance and correlation among samples, and effectively reduces the dimensionality of data while preserving the spatial geometric structure.
IEEE TRANSACTIONS ON RELIABILITY
(2023)
Article
Computer Science, Artificial Intelligence
Bolin Wang, Yuanyuan Sun, Yonghe Chu, Zhihao Yang, Hongfei Lin
Summary: In this study, a global-locality preserving projection method is proposed to refine word representation, by re-embedding word vectors to a manifold semantic space. It extracts local and global features of word vectors, discovers latent semantic structure, and obtains a compact word embedding subspace.
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2022)
Article
Computer Science, Artificial Intelligence
Ruisheng Ran, Hao Qin, Shougui Zhang, Bin Fang
Summary: A simple and robust LPP method, LPPMDC, is proposed in this paper to solve the small-sample-size problem and improve the performance stability of LPP under varying neighborhood size parameter. Experiment results on three face databases demonstrate the efficiency and robustness of LPPMDC.
NEURAL PROCESSING LETTERS
(2022)
Article
Computer Science, Artificial Intelligence
Mingjie He, Jie Zhang, Shiguang Shan, Xiao Liu, Zhongqin Wu, Xilin Chen
Summary: In this paper, a novel method for simulating occlusion by dropping the activations of a group of neurons is proposed, along with an attention module to improve the contributions of non-occluded regions. Experimental results show that the proposed method achieves significant improvements in the robustness and accuracy of face recognition.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2022)
Article
Computer Science, Artificial Intelligence
Jiayan Qiu, Xinchao Wang, Pascal Fua, Dacheng Tao
Summary: In this paper, a novel unsupervised approach for sequence matching is proposed, using sequencelet to match sequences with strong similarities and grouping frames together. The optimal sequencelets and matching between them are learned jointly, without supervision. The method outperforms state-of-the-art ones on datasets of different domains.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2021)
Article
Computer Science, Information Systems
Zixun Xiong, Minghua Wan, Rui Xue, Guowei Yang
Summary: This paper introduces an improved algorithm called 2D-MELPP, a novel matrix exponential method to enhance the performance of 2D-LPP. By replacing some matrix multiplications with multiple multiplications, it provides an efficient way for solving 2D-MELPP. Experimental results show that 2D-MELPP outperforms other methods in recognition accuracy on multiple public databases.
KSII TRANSACTIONS ON INTERNET AND INFORMATION SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Yixiong Liang, Zhihong Tang, Meng Yan, Jialin Chen, Qing Liu, Yao Xiang
Summary: Automated detection of cervical cancer cells/clumps has the potential to significantly reduce error rate and increase productivity in cervical cancer screening. The proposed Comparison detector shows improvements in detection efficiency and accuracy compared to traditional methods.
Article
Microscopy
Yao Xiang, Zhujun He, Qing Liu, Jialin Chen, Yixiong Liang
Summary: The proposed autofocus algorithm for whole slide imaging utilizes convolution and recurrent neural networks to predict the out-of-focus distance and update the focus location of the camera lens iteratively, achieving rapid determination of the optimal in-focus image.
Article
Computer Science, Interdisciplinary Applications
Yixiong Liang, Changli Pan, Wanxin Sun, Qing Liu, Yun Du
Summary: This study proposes a global context-aware framework to address the issue of too many false positive predictions in computer-aided cervical cancer screening, by integrating global context information through image-level classification branch and weighted loss. Additionally, a new ground truth assignment strategy called soft scale anchor matching is introduced to facilitate feature learning by softly matching ground truths with anchors across scales.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2021)
Article
Computer Science, Information Systems
Qing Liu, Haotian Liu, Yang Zhao, Yixiong Liang
Summary: This paper proposes a dual-branch network with dual-sampling modulated Dice loss for automated segmentation of hard exudates in colour fundus images. The method tackles the challenges of extreme class imbalance and enormous size variation by utilizing two branches and different sampling strategies.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2022)
Article
Robotics
Ping Zhong, Bolei Chen, Siyi Lu, Xiaoxi Meng, Yixiong Liang
Summary: This paper proposes an information-driven exploration strategy for Unmanned Aerial Vehicles (UAVs) in unknown environments, utilizing the fast marching method. The strategy includes frontier point detection, evaluation of candidate goals based on a utility function, and optimization of UAV trajectory and yaw angle. Simulation experiments demonstrate the superiority of the proposed strategy.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Computer Science, Artificial Intelligence
Qing Liu, Haotian Liu, Wei Ke, Yixiong Liang
Summary: This paper proposes a novel Many-to-Many Reassembly of Features (M2MRF) for tiny lesion segmentation, which can reassemble features in a dimension-reduced feature space and simultaneously aggregates multiple features inside a large predefined region. Experimental results show that our M2MRF outperforms existing feature reassembly operators and achieves better performances and generalization ability than existing methods when combined with HRNetV2.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Artificial Intelligence
Tao Sheng, Chengchao Shen, Yuan Liu, Yeyu Ou, Zhe Qu, Yixiong Liang, Jianxin Wang
Summary: Federated learning connects decentralized datasources for joint training of deep models, reducing the risk of privacy leakage. However, label distribution skew can impair the performance of the global model. To address this, FedMGD proposes a novel method using a global Generative Adversarial Network to model the data distribution without accessing local datasets, resulting in improved performance without privacy concerns. Experimental results demonstrate its superiority over state-of-the-art methods. Code is available at https://www.github.com/Sheng-T/FedMGD.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Information Systems
Yixiong Liang, Shuo Feng, Qing Liu, Hulin Kuang, Jianfeng Liu, Liyan Liao, Yun Du, Jianxin Wang
Summary: To improve the performance of cervical abnormal cell detection, we propose a method that utilizes contextual relationships. By exploring the relationships between cells and cell-to-global images, the features of each region of interest are enhanced. Our experiments on a large dataset validate the effectiveness of the proposed method by achieving better average precision (AP) than baseline methods.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Computer Science, Information Systems
Hulin Kuang, Yahui Wang, Yixiong Liang, Jin Liu, Jianxin Wang
Summary: This study proposes a new body and edge aware network for medical image segmentation. It introduces various modules and applies deep supervision to effectively extract body and edge features, resulting in improved segmentation performance.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Biochemical Research Methods
Yixiong Liang, Zhihua Yin, Haotian Liu, Hailong Zeng, Jianxin Wang, Jianfeng Liu, Nanying Che
Summary: This paper proposes a novel weakly supervised nuclei segmentation framework that utilizes sparsely annotated bounding boxes for segmentation without requiring segmentation labels. The framework integrates traditional image segmentation and self-training to achieve fully supervised instance segmentation. The method generates coarse masks and pseudo labels using the traditional segmentation and refines them with the help of a teacher model. Both teacher and student models are trained jointly using the refined masks, pseudo labels, and manually annotated bounding boxes. The proposed method outperforms existing weakly supervised methods on both datasets.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2023)
Article
Automation & Control Systems
Bolei Chen, Siyi Lu, Ping Zhong, Yongzheng Cui, Yixiong Liang, Jianxin Wang
Summary: This paper proposes a new strategy for target-driven semantic navigation by considering ternary feature fusion of human-robot-object. Through the integration of multi-granularity map features and social awareness, and the use of deep reinforcement learning and dual-channel value estimation network, the challenges of dynamic and crowded scenarios in TDSN are effectively addressed.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2024)
Proceedings Paper
Acoustics
Yongsheng Zhang, Qing Liu, Yang Zhao, Yixiong Liang
Summary: Unsupervised visual representation learning aims to learn general features from unlabelled data. We propose a new method with jigsaw clustering and classification as pretext tasks to achieve competitive results to contrastive learning with low computational overhead.
2022 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP)
(2022)
Proceedings Paper
Acoustics
Ruixiang Geng, Qing Liu, Shuo Feng, Yixiong Liang
Summary: This study proposes a two-stage learning framework for fully automated cervical cancer grading. It first learns deep pathological features at the patch level through a cell instance detection task, and then uses these patch-level features to learn the WSI-level pathological features for cervical cancer grading. Experimental results demonstrate that this method achieves state-of-the-art performance in cervical cancer grading.
2022 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP)
(2022)
Proceedings Paper
Imaging Science & Photographic Technology
Yao Xiang, Jialin Chen, Qing Liu, Yixiong Liang
2020 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP)
(2020)
Proceedings Paper
Acoustics
Meng Yan, Qing Liu, Zhihua Yin, Du Wang, Yixiong Liang
2020 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING
(2020)
Article
Computer Science, Artificial Intelligence
Rui Lv, Dingheng Wang, Jiangbin Zheng, Zhao-Xu Yang
Summary: In this paper, the authors investigate tensor decomposition for neural network compression. They analyze the convergence and precision of tensor mapping theory, validate the rationality of tensor mapping and its superiority over traditional tensor approximation based on the Lottery Ticket Hypothesis. They propose an efficient method called 3D-KCPNet to compress 3D convolutional neural networks using the Kronecker canonical polyadic (KCP) tensor decomposition. Experimental results show that 3D-KCPNet achieves higher accuracy compared to the original baseline model and the corresponding tensor approximation model.
Article
Computer Science, Artificial Intelligence
Xiangkun He, Zhongxu Hu, Haohan Yang, Chen Lv
Summary: In this paper, a novel constrained multi-objective reinforcement learning algorithm is proposed for personalized end-to-end robotic control with continuous actions. The approach trains a single model using constraint design and a comprehensive index to achieve optimal policies based on user-specified preferences.
Article
Computer Science, Artificial Intelligence
Zhijian Zhuo, Bilian Chen, Shenbao Yu, Langcai Cao
Summary: In this paper, a novel method called Expansion with Contraction Method for Overlapping Community Detection (ECOCD) is proposed, which utilizes non-negative matrix factorization to obtain disjoint communities and applies expansion and contraction processes to adjust the degree of overlap. ECOCD is applicable to various networks with different properties and achieves high-quality overlapping community detection.
Article
Computer Science, Artificial Intelligence
Yizhe Zhu, Chunhui Zhang, Jialin Gao, Xin Sun, Zihan Rui, Xi Zhou
Summary: In this work, the authors propose a Contrastive Spatio-Temporal Distilling (CSTD) approach to improve the detection of high-compressed deepfake videos. The approach leverages spatial-frequency cues and temporal-contrastive alignment to fully exploit spatiotemporal inconsistency information.
Review
Computer Science, Artificial Intelligence
Laijin Meng, Xinghao Jiang, Tanfeng Sun
Summary: This paper provides a review of coverless steganographic algorithms, including the development process, known contributions, and general issues in image and video algorithms. It also discusses the security of coverless steganography from theoretical analysis to actual investigation for the first time.
Article
Computer Science, Artificial Intelligence
Yajie Bao, Tianwei Xing, Xun Chen
Summary: Visual question answering requires processing multi-modal information and effective reasoning. Neural-symbolic learning is a promising method, but current approaches lack uncertainty handling and can only provide a single answer. To address this, we propose a confidence based neural-symbolic approach that evaluates NN inferences and conducts reasoning based on confidence.
Article
Computer Science, Artificial Intelligence
Anh H. Vo, Bao T. Nguyen
Summary: Interior style classification is an interesting problem with potential applications in both commercial and academic domains. This project proposes a method named ISC-DeIT, which combines data-efficient image transformer architectures and knowledge distillation, to address the interior style classification problem. Experimental results demonstrate a significant improvement in predictive accuracy compared to other state-of-the-art methods.
Article
Computer Science, Artificial Intelligence
Shashank Kotyan, Danilo Vasconcellos Vargas
Summary: This article introduces a novel augmentation technique called Dynamic Scanning Augmentation to improve the accuracy and robustness of Vision Transformer (ViT). The technique leverages dynamic input sequences to adaptively focus on different patches, resulting in significant changes in ViT's attention mechanism. Experimental results demonstrate that Dynamic Scanning Augmentation outperforms ViT in terms of both robustness to adversarial attacks and accuracy against natural images.
Article
Computer Science, Artificial Intelligence
Hiba Alqasir, Damien Muselet, Christophe Ducottet
Summary: The article proposes a solution to improve the learning process of a classification network by providing shape priors, reducing the need for annotated data. The solution is tested on cross-domain digit classification tasks and a video surveillance application.
Article
Computer Science, Artificial Intelligence
Dexiu Ma, Mei Liu, Mingsheng Shang
Summary: This paper proposes a method using neural dynamics solvers to solve infinity-norm optimization problems. Two improved solvers are constructed and their effectiveness and superiority are demonstrated through theoretical analysis and simulation experiments.
Article
Computer Science, Artificial Intelligence
Francesco Gregoretti, Giovanni Pezzulo, Domenico Maisto
Summary: Active Inference is a computational framework that uses probabilistic inference and variational free energy minimization to describe perception, planning, and action. cpp-AIF is a header-only C++ library that provides a powerful tool for implementing Active Inference for Partially Observable Markov Decision Processes through multi-core computing. It is cross-platform and improves performance, memory management, and usability compared to existing software.
Article
Computer Science, Artificial Intelligence
Zelin Ying, Dawei Cheng, Cen Chen, Xiang Li, Peng Zhu, Yifeng Luo, Yuqi Liang
Summary: This paper proposes a novel stock market trends prediction framework called SMART, which includes a self-supervised stock technical data sequence embedding model S3E. By training with multiple self-supervised auxiliary tasks, the model encodes stock technical data sequences into embeddings and uses the learned sequence embeddings for predicting stock market trends. Extensive experiments on China A-Shares market and NASDAQ market prove the high effectiveness of our model in stock market trends prediction, and its effectiveness is further validated in real-world applications in a leading financial service provider in China.
Article
Computer Science, Artificial Intelligence
Hao Li, Hao Jiang, Dongsheng Ye, Qiang Wang, Liang Du, Yuanyuan Zeng, Liu Yuan, Yingxue Wang, C. Chen
Summary: DHGAT1, a dynamic hyperbolic graph attention network, utilizes hyperbolic metric properties to embed dynamic graphs. It employs a spatiotemporal self-attention mechanism and weighted node representations, resulting in excellent performance in link prediction tasks.
Article
Computer Science, Artificial Intelligence
Jiehui Huang, Zhenchao Tang, Xuedong He, Jun Zhou, Defeng Zhou, Calvin Yu-Chian Chen
Summary: This study proposes a progressive learning multi-scale feature blending model for image deraining tasks. The model utilizes detail dilation and texture extraction to improve the restoration of rainy images. Experimental results show that the model achieves near state-of-the-art performance in rain removal tasks and exhibits better rain removal realism.
Article
Computer Science, Artificial Intelligence
Lizhi Liu, Zilin Gao, Yinhe Wang, Yongfu Li
Summary: This paper proposes a novel discrete-time interconnected model for depicting complex dynamical networks. The model consists of nodes and edges subsystems, which consider the dynamic characteristic of both nodes and edges. By designing control strategies and coupling modes, the stabilization and synchronization of the network are achieved. Simulation results demonstrate the effectiveness of the proposed methods.