Article
Environmental Sciences
Usman Muhammad, Md Ziaul Hoque, Weiqiang Wang, Mourad Oussalah
Summary: This study proposes a patch-based discriminative learning (PBDL) method for remote sensing scene classification. The method enhances the discriminative power of image representation by employing multi-level feature learning and image pyramids. Experimental results demonstrate that the proposed method outperforms existing deep-learning-based methods in classification performance.
Article
Engineering, Multidisciplinary
Chaowei Lin, Feifei Lee, Jiawei Cai, Hanqing Chen, Qiu Chen
Summary: The research introduces a novel scene recognition framework, combining graph encoded local discriminative region representation and multi-head attention module to enhance model performance.
CMES-COMPUTER MODELING IN ENGINEERING & SCIENCES
(2021)
Article
Computer Science, Software Engineering
Zongji Wang, Yunfei Liu, Feng Lu
Summary: This paper utilizes deep learning to solve the challenging computer vision problem of intrinsic image decomposition, achieving high efficiency by extracting discriminative features for different intrinsic layers. Experimental results demonstrate that the proposed network structure outperforms existing state-of-the-art methods.
COMPUTATIONAL VISUAL MEDIA
(2023)
Article
Computer Science, Information Systems
Haitao Zeng, Xinhang Song, Gongwei Chen, Shuqiang Jiang
Summary: This paper proposes a novel scene recognition framework that detects discriminative regions with non-geometric contours surrounding targets tightly and embeds both foreground and background regions with a graph model to obtain scene representations. Experimental results demonstrate the effectiveness and generality of the proposed method on MIT67 and SUN397 datasets.
IEEE TRANSACTIONS ON MULTIMEDIA
(2022)
Article
Computer Science, Artificial Intelligence
Gregg Belous, Andrew Busch, Yongsheng Gao
Summary: The DSDPL framework is proposed for multi-category image classification, decomposing data into class-shared and class-specific subspaces via learned projection matrices for a more stable classification approach.
PATTERN RECOGNITION
(2021)
Article
Computer Science, Information Systems
Jie Mei, Wei Liu, Ming Zhu, Yongka Qi, Ming Fu, Yushi Li, Quan Yuan
Summary: This study investigates the discriminative angle deep features for vibration signals and proposes a novel normalized one-versus-all classification loss with center and contrastive regularization. The trained network optimizes deep features to ensure intra-class compactness and inter-class divergence, and can be used for open-set fault classification. Furthermore, the proposed method is verified using field-measured motor bearing and gear vibration signals, showing evident advantages over other approaches in practical fault diagnosis scenarios.
Article
Engineering, Electrical & Electronic
Fan Wang, Dong Yin
Summary: Most scene text recognition methods use synthetic data to train models, but there is a gap between synthetic and real data. To solve this problem, we trained a new model using a few real data and achieved improved accuracy through various techniques.
JOURNAL OF ELECTRONIC IMAGING
(2022)
Article
Geochemistry & Geophysics
Jue Wang, He Chen, Long Ma, Liang Chen, Xiaodong Gong, Wenchao Liu
Summary: The paper proposes a novel deep metric learning loss function for feature extraction in remote sensing scene classification. By introducing geometric and spatial constraints, this loss function enhances the discrimination of feature representations, leading to improved classification performance.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Computer Science, Artificial Intelligence
Tobias Schlagenhauf, Yiwen Lin, Benjamin Noack
Summary: This work proposes a novel method that forces a set of base models to learn different features for a classification task, which can improve the classification accuracy. Experimental results demonstrate the effectiveness of this method.
MACHINE VISION AND APPLICATIONS
(2023)
Article
Ecology
Zhengyu Zhao, Yuanyuan Lu, Yijie Tong, Xin Chen, Ming Bai
Summary: Digitalized natural history collections are important resources for ecological and evolutionary research. Our PENet model can automatically extract and encode features of specimens, enabling efficient specimen retrieval using hash codes.
METHODS IN ECOLOGY AND EVOLUTION
(2023)
Article
Computer Science, Artificial Intelligence
Zhitong Xiong, Yuan Yuan, Qi Wang
Summary: An efficient framework for RGB-D scene recognition is proposed in this article, which adaptively selects important local features to capture the spatial variability of scene images. By designing a differentiable local feature selection (DLFS) module, key local scene-related features can be extracted from spatially-correlated multi-modal RGB-D features. By concatenating local-orderless and global-structured multi-modal features, the proposed framework achieves state-of-the-art performance on public RGB-D scene recognition datasets.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2021)
Article
Computer Science, Artificial Intelligence
Chen Wang, Guohua Peng, Bernard De Baets
Summary: In this paper, a novel class-specific discriminative metric learning method (CSDML) is proposed to address the problems caused by high within-class diversity and high between-class similarity in scene recognition. By learning a distinctive linear transformation or Mahalanobis distance metric for each class, the proposed method projects samples into a low-dimensional discriminative space and achieves better metric learning performance by minimizing the distances between the projections of samples from the same class and maximizing the distances between the projections of samples from different classes. Experimental results demonstrate the superiority of the proposed method over existing approaches on four benchmark scene datasets.
PATTERN RECOGNITION
(2022)
Article
Computer Science, Information Systems
Tuya
Summary: This paper proposes a graph convolutional enhanced discriminative broad learning system (GCDBLS) for hyperspectral image (HSI) classification, which aggregates node information through graph convolution and introduces the concept of local scatter to extract more discriminative features, achieving good results in HSI classification.
Article
Remote Sensing
Junjie Zhu, Ke Yang, Naiyang Guan, Xiaodong Yi, Chunping Qiu
Summary: In this study, a novel and effective approach for few-shot remote sensing scene classification is proposed. The approach incorporates the query feature as a key factor in the prototype formation and enhances the discriminativeness of the prototypes. Experimental results demonstrate that the proposed method outperforms the state-of-the-art methods on popular datasets.
INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATION AND GEOINFORMATION
(2023)
Article
Computer Science, Artificial Intelligence
Bao-Qing Yang, Xin-Ping Guan, Jun-Wu Zhu, Chao-Chen Gu, Kai-Jie Wu, Jia-Jie Xu
Summary: The paper presents a discriminative dictionary learning framework based on support vector machines and feedback mechanism to enhance image classification performance.
PATTERN RECOGNITION
(2021)
Article
Computer Science, Information Systems
Lin Xie, Feifei Lee, Li Liu, Zhong Yin, Qiu Chen
IEEE TRANSACTIONS ON MULTIMEDIA
(2020)
Article
Chemistry, Analytical
Hanqing Chen, Chunyan Hu, Feifei Lee, Chaowei Lin, Wei Yao, Lu Chen, Qiu Chen
Summary: This study introduces a content-based video retrieval system using the deep supervised video hashing (DSVH) framework, and demonstrates its advantages through experiments on different datasets.
Article
Computer Science, Artificial Intelligence
Qian Zhang, Feifei Lee, Ya-gang Wang, Damin Ding, Wei Yao, Lu Chen, Qiu Chen
Summary: The paper introduces a novel end-to-end framework for noise correction, which can completely correct noisy labels to true labels and keep the number of each class more balanced without requiring any extra conditions. Experimental results show that the proposed method outperforms other state-of-the-art methods on publicly available CIFAR-10, CIFAR-100 and Clothing1M datasets.
APPLIED SOFT COMPUTING
(2021)
Article
Computer Science, Information Systems
Qian Zhang, Feifei Lee, Ya-gang Wang, Damin Ding, Shuai Yang, Chaowei Lin, Qiu Chen
Summary: The paper introduces a novel framework CJC-net for learning with noisy labels, which utilizes cyclical training with joint loss and co-teaching strategy to help networks transition from overfitting to underfitting states, thereby improving the accuracy in identifying noisy labels and hard samples.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Artificial Intelligence
Chaowei Lin, Feifei Lee, Lin Xie, Jiawei Cai, Hanqing Chen, Li Liu, Qiu Chen
Summary: In this paper, a comprehensive representation for scene recognition is proposed, which includes enhanced global scene representation, local salient scene representation, and local contextual object representation. The multiple representations are constructed using two pretrained CNNs and specific techniques, and they are generated by an end-to-end trainable model. Experimental results show that the proposed model outperforms existing models.
APPLIED SOFT COMPUTING
(2022)
Article
Computer Science, Interdisciplinary Applications
Bin Zuo, Feifei Lee, Qiu Chen
Summary: Skin lesion segmentation is crucial in skin diagnosis. This paper proposes a novel U-shaped network called EAM-CPFNet, which combines edge attention module (EAM) and context pyramid fusion (CPF) to improve the performance of skin lesion segmentation. Experimental results show that the proposed method is competitive on a publicly available dataset.
MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING
(2022)
Article
Computer Science, Artificial Intelligence
Zhihong Yu, Feifei Lee, Qiu Chen
Summary: In this paper, a hybrid CNN-Transformer model based on a neural architecture search network (HCT-Net) is proposed, which can capture feature information, save time and human energy, and reduce GPU memory consumption and complexity. The model achieves competitive precision and efficiency on various medical image segmentation datasets, and its generalization is validated on unseen datasets.
APPLIED INTELLIGENCE
(2023)
Article
Chemistry, Multidisciplinary
Xijun Xie, Feifei Lee, Qiu Chen
Summary: This paper proposes a novel framework called DMA-Net, which utilizes the DMAM module to perform multi-scale feature extraction and information fusion, and the DGM module to reduce the impact of information exchange between branches. DMA-Net achieves incremental FSOD and demonstrates state-of-the-art performance in this setting.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Information Systems
Qiangqiang Xia, Feifei Lee, Qiu Chen
Summary: With the development of deep neural networks, there is a growing demand for accurately labeled datasets. However, human-labeled datasets often have mistakes, leading to misleading information. This paper proposes a two-stage learning framework called TCC-net to address the issue of learning with noisy labels. The experimental results show that TCC-net outperforms other state-of-the-art methods on corrupted datasets.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Information Systems
Da-Min Ding, Ya-Gang Wang, Wei Zhang, Qiu Chen
Summary: This paper proposes an improved fall detection method on a smart walker based on IMU sensors and image recognition, which can improve detection accuracy and real-time performance, making it an ideal solution.
Article
Computer Science, Information Systems
Zhiyi Feng, Feifei Lee, Qiu Chen
Summary: In recent years, lightweight models have been successfully applied to single image super-resolution tasks, but most models fail to fully utilize multi-scale features. To address this issue, we propose a stacked reversed U-shape network (SRUNet) that progressively performs upsampling and downsampling operations to extract richer multi-scale features. Additionally, we introduce dense connections and fusion modules for better utilization of multi-scale features.
Article
Computer Science, Information Systems
Hui Wei, Feifei Lee, Chunyan Hu, Qiu Chen
Summary: This paper proposes an efficient CNN architecture search framework, MOO-DNAS, based on multi-objective optimization. The framework aims to find an efficient model by balancing classification accuracy and network latency, utilizing a novel factorized hierarchical search space and a robust hard-sampling strategy.
Article
Computer Science, Information Systems
Zhenwen Cai, Feifei Lee, Chunyan Hu, Koji Kotani, Qiu Chen
Summary: The paper introduces a novel lightweight and general neural network module NAEM, which achieves significant performance improvement in deep reinforcement learning by introducing Gaussian noise into the attention mechanism for global exploration.
Article
Computer Science, Information Systems
Wei Wang, Feifei Lee, Shuai Yang, Qiu Chen
Summary: CFR-CapsNet is an improved capsule network that uses CFR and self-attention mechanism to enhance CapsNet performance, and improves network structure relevance through information transmission. Experimental results show that this method can effectively improve the performance of CapsNet.
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)