NSTL国家科技图书文献中心

1. LDINet: Long Distance Imaging Through RGB and NIR Image Fusion NSTL国家科技图书文献中心

Lin Mei | Hao Zhang ... - 《Pattern Recognition,Part XXVI》 - International Conference on Pattern Recognition - 2025, - 171～185 - 共15页

摘要： adopt pyramid* feature selection to capture multiscale* | The fusion of visible color (RGB) and near | infrared (NIR) images takes multispectral advantage of | colors from RGB image and details from NIR image | . Unlike RGB images, NIR images are robust to atmospheric

关键词： Image fusion | Attention map | Long distance imaging | Near infrared | Pyramid feature selection

2. Rethinking Features-Fused-Pyramid-Neck for Object Detection NSTL国家科技图书文献中心

Hulin Li - 《Computer Vision - ECCV 2024,Part LXVII》 - European Conference on Computer Vision - 2025, - 74～90 - 共17页

摘要：. Furthermore, we present a feature* adaptive selection method* | features-fused-pyramid-neck for multi-scale detection and | approach faces feature* misalignment when representations* | from different hierarchical levels of the feature | pyramid are forcibly fused point-to-point. To address

关键词： Object detection | Feature pyramid | Feature misalignment | Detection-net architecture

3. FsPN: Blind Image Quality Assessment Based on Feature-Selected Pyramid Network NSTL国家科技图书文献中心

Long Tang | Yongming Han ... - 《IEEE signal processing letters》 - 2025,32 - 1～5 - 共5页

摘要： and then perform feature* fusion. However, the* | -train the feature* extraction network from scratch* | resource demands. In this letter, a Feature-selected | Pyramid Network(FsPN) is proposed to address this issue | selection module selects useful information from the

关键词： Feature extraction | Image quality | Convolution | Distortion | Residual neural networks | Computational modeling | Interpolation | Filters | Transformers | Training

4. Multi-modal Dynamic Information Selection Pyramid Network for Alzheimer's Disease Classification NSTL国家科技图书文献中心

Yuanmin Ma | Yuan Chen ... - 《Advances in Brain Inspired Cognitive Systems,Part II》 - International Conference on Advances in Brain Inspired Cognitive Systems - 2025, - 1～10 - 共10页

摘要： feature pyramid for finer detail extraction and a cross | -modal dynamic information selection* framework to* | three main modules: a dual ResNet50-based feature | dynamic information selection* module that evaluates the* | Alzheimer's disease (AD) is a common and

关键词： Brain disease prediction | Alzheimer's disease | Multi-modal classification

5. CAFS-Net: Context-Aware Feature Selection Network for Accurate and Efficient Tiny Surface Defect Segmentation NSTL国家科技图书文献中心

Tian, Zhonghua | Yang, Xianqiang - 《IEEE Transactions on Instrumentation and Measurement》 - 2025,74(Pt.1) - 5013011.1～5013011.11 - 共11页

摘要： propose a context-aware feature* selection network (CAFS* | network (CANet), a novel feature* selection module (FSM* | ), and a lightweight feature-fusing decoder. The CANet | defective features of image blocks from the feature | pyramid of image blocks based on the classification

关键词： Feature extraction | Transformers | Semantic segmentation | Inspection | Training | Accuracy | Noise | Decoding | Computer vision | Computational modeling...

6. A Lightweight Network for Ship Detection in SAR Images Based on Edge Feature Aware and Fusion NSTL国家科技图书文献中心

Yuming Li | Jin Liu ... - 《IEEE journal of selected topics in applied earth observations and remote sensing》 - 2025,18 - 3782～3796 - 共15页

摘要： network, Filter-Pruned Bi-directional Feature Pyramid | lightweight detection method based on edge feature* aware and* | feature, we introduce an Edge Feature-Aware (EFA | module. Furthermore, a lightweight feature* fusion* | quantization algorithm based on a bit-width selection

关键词： Marine vehicles | Feature extraction | Image edge detection | Radar polarimetry | Accuracy | Quantization (signal) | Synthetic aperture radar | Computational complexity | Adaptation models | Inference algorithms

7. A multiscale approach for network intrusion detection based on variance-covariance subspace distance and EQL v2 NSTL国家科技图书文献中心

Taotao Liu | Yu Fu ... - 《Computers & Security》 - 2025,148(Jan.) - 104173.1～104173.15 - 共15页

摘要： distance is used for feature* selection on the* | feature redundancy and class imbalance problem in the | feature subsets that can effectively approximate the | original feature* space. Secondly, the loss function, EQL* | , a pyramid* depthwise separable convolution model is*

关键词： Intrusion detection | Feature selection | Class imbalance | Multiscale modeling

8. PDSMV3-DCRNN: A novel ensemble deep learning framework for enhancing phishing detection and URL extraction NSTL国家科技图书文献中心

Y. Bhanu Prasad | Venkatesulu Dondeti - 《Computers & Security》 - 2025,148(Jan.) - 104123.1～104123.16 - 共16页

摘要： feature selection process, which includes the | feature selection. This dynamic approach enhances the | methods rely heavily on predefined feature sets, which | , resulting in a feature-reduced dataset used for training | , specifically the Novel Pyramid* Depth-wise Separable*

关键词： Phishing | Uniform resources locator (URL) | Binary grey goose optimization algorithm | (BGGOA) | Pyramid depth wise separable-mobileNetV3 | (PyDS-MV3) | Deformable convolutional residual neural | network (DCRNN) | Boosted ConvNeXt

9. Fast Visual-Inertial Odometry with Adaptive Feature Coupling NSTL国家科技图书文献中心

Zekun Ma | Jiazheng Xiao ... - 《Intelligent Robotics and Applications,Part VI》 - International Conference on Intelligent Robotics and Applications - 2025, - 171～186 - 共16页

摘要：, which use concise head feature* pooling pyramid and* | adaptive feature* selection gate is introduced to reduce* | cost and robust feature* coupling mechanism. However* | In recent years, visual-inertial odometry (VIO | ) has been widely applied in localization, navigation

关键词： Visual-Inertial odometry | Pose estimation | Multimodal feature fusion | Adaptive learning

10. Advanced Image Enhancement and a Lightweight Feature Pyramid Network for Detecting Microaneurysms in Diabetic Retinopathy Screening NSTL国家科技图书文献中心

Muhammad Zeeshan Tah... | Xingzheng Lyu ... - 《International journal of imaging systems and technology》 - 2025,35(1) - e70004.1～e70004.15 - 共15页

摘要： lightweight feature* pyramid network (FPN) with a pretrained* | convolutional block attention module (CBAM) to enhance feature | selection. CBAM applies spatial and channel-wise attention | Diabetic retinopathy (DR) is a complication of | diabetes that can lead to vision impairment and even

摘要： adopt pyramid* feature selection to capture multiscale* | The fusion of visible color (RGB) and near | infrared (NIR) images takes multispectral advantage of | colors from RGB image and details from NIR image | . Unlike RGB images, NIR images are robust to atmospheric

关键词： Image fusion | Attention map | Long distance imaging | Near infrared | Pyramid feature selection

关键词： Object detection | Feature pyramid | Feature misalignment | Detection-net architecture

摘要： and then perform feature* fusion. However, the* | -train the feature* extraction network from scratch* | resource demands. In this letter, a Feature-selected | Pyramid Network(FsPN) is proposed to address this issue | selection module selects useful information from the

关键词： Feature extraction | Image quality | Convolution | Distortion | Residual neural networks | Computational modeling | Interpolation | Filters | Transformers | Training

摘要： feature pyramid for finer detail extraction and a cross | -modal dynamic information selection* framework to* | three main modules: a dual ResNet50-based feature | dynamic information selection* module that evaluates the* | Alzheimer's disease (AD) is a common and

关键词： Brain disease prediction | Alzheimer's disease | Multi-modal classification

摘要： propose a context-aware feature* selection network (CAFS* | network (CANet), a novel feature* selection module (FSM* | ), and a lightweight feature-fusing decoder. The CANet | defective features of image blocks from the feature | pyramid of image blocks based on the classification

关键词： Feature extraction | Transformers | Semantic segmentation | Inspection | Training | Accuracy | Noise | Decoding | Computer vision | Computational modeling...

摘要： network, Filter-Pruned Bi-directional Feature Pyramid | lightweight detection method based on edge feature* aware and* | feature, we introduce an Edge Feature-Aware (EFA | module. Furthermore, a lightweight feature* fusion* | quantization algorithm based on a bit-width selection

关键词： Marine vehicles | Feature extraction | Image edge detection | Radar polarimetry | Accuracy | Quantization (signal) | Synthetic aperture radar | Computational complexity | Adaptation models | Inference algorithms

摘要： distance is used for feature* selection on the* | feature redundancy and class imbalance problem in the | feature subsets that can effectively approximate the | original feature* space. Secondly, the loss function, EQL* | , a pyramid* depthwise separable convolution model is*

关键词： Intrusion detection | Feature selection | Class imbalance | Multiscale modeling

摘要： feature selection process, which includes the | feature selection. This dynamic approach enhances the | methods rely heavily on predefined feature sets, which | , resulting in a feature-reduced dataset used for training | , specifically the Novel Pyramid* Depth-wise Separable*

关键词： Phishing | Uniform resources locator (URL) | Binary grey goose optimization algorithm | (BGGOA) | Pyramid depth wise separable-mobileNetV3 | (PyDS-MV3) | Deformable convolutional residual neural | network (DCRNN) | Boosted ConvNeXt

摘要：, which use concise head feature* pooling pyramid and* | adaptive feature* selection gate is introduced to reduce* | cost and robust feature* coupling mechanism. However* | In recent years, visual-inertial odometry (VIO | ) has been widely applied in localization, navigation

关键词： Visual-Inertial odometry | Pose estimation | Multimodal feature fusion | Adaptive learning

关键词： diabetic retinopathy | feature pyramid network | image enhancement | microaneurysm detection

4008-161-200 800-990-8900

国家科技图书文献中心

摘要： adopt pyramid feature selection to capture multiscale | The fusion of visible color (RGB) and near | infrared (NIR) images takes multispectral advantage of | colors from RGB image and details from NIR image | . Unlike RGB images, NIR images are robust to atmospheric

关键词： Image fusion | Attention map | Long distance imaging | Near infrared | Pyramid feature selection

摘要：. Furthermore, we present a feature adaptive selection method | features-fused-pyramid-neck for multi-scale detection and | approach faces feature misalignment when representations | from different hierarchical levels of the feature | pyramid are forcibly fused point-to-point. To address

关键词： Object detection | Feature pyramid | Feature misalignment | Detection-net architecture

摘要： and then perform feature fusion. However, the | -train the feature extraction network from scratch | resource demands. In this letter, a Feature-selected | Pyramid Network(FsPN) is proposed to address this issue | selection module selects useful information from the

关键词： Feature extraction | Image quality | Convolution | Distortion | Residual neural networks | Computational modeling | Interpolation | Filters | Transformers | Training

摘要： feature pyramid for finer detail extraction and a cross | -modal dynamic information selection framework to | three main modules: a dual ResNet50-based feature | dynamic information selection module that evaluates the | Alzheimer's disease (AD) is a common and

关键词： Brain disease prediction | Alzheimer's disease | Multi-modal classification

摘要： propose a context-aware feature selection network (CAFS | network (CANet), a novel feature selection module (FSM | ), and a lightweight feature-fusing decoder. The CANet | defective features of image blocks from the feature | pyramid of image blocks based on the classification

关键词： Feature extraction | Transformers | Semantic segmentation | Inspection | Training | Accuracy | Noise | Decoding | Computer vision | Computational modeling...

摘要： network, Filter-Pruned Bi-directional Feature Pyramid | lightweight detection method based on edge feature aware and | feature, we introduce an Edge Feature-Aware (EFA | module. Furthermore, a lightweight feature fusion | quantization algorithm based on a bit-width selection

关键词： Marine vehicles | Feature extraction | Image edge detection | Radar polarimetry | Accuracy | Quantization (signal) | Synthetic aperture radar | Computational complexity | Adaptation models | Inference algorithms

摘要： distance is used for feature selection on the | feature redundancy and class imbalance problem in the | feature subsets that can effectively approximate the | original feature space. Secondly, the loss function, EQL | , a pyramid depthwise separable convolution model is

关键词： Intrusion detection | Feature selection | Class imbalance | Multiscale modeling

摘要： feature selection process, which includes the | feature selection. This dynamic approach enhances the | methods rely heavily on predefined feature sets, which | , resulting in a feature-reduced dataset used for training | , specifically the Novel Pyramid Depth-wise Separable

关键词： Phishing | Uniform resources locator (URL) | Binary grey goose optimization algorithm | (BGGOA) | Pyramid depth wise separable-mobileNetV3 | (PyDS-MV3) | Deformable convolutional residual neural | network (DCRNN) | Boosted ConvNeXt

摘要：, which use concise head feature pooling pyramid and | adaptive feature selection gate is introduced to reduce | cost and robust feature coupling mechanism. However | In recent years, visual-inertial odometry (VIO | ) has been widely applied in localization, navigation

关键词： Visual-Inertial odometry | Pose estimation | Multimodal feature fusion | Adaptive learning

摘要： lightweight feature pyramid network (FPN) with a pretrained | convolutional block attention module (CBAM) to enhance feature | selection. CBAM applies spatial and channel-wise attention | Diabetic retinopathy (DR) is a complication of | diabetes that can lead to vision impairment and even

关键词： diabetic retinopathy | feature pyramid network | image enhancement | microaneurysm detection

4008-161-200 800-990-8900

国家科技图书文献中心

摘要： adopt pyramid* feature selection to capture multiscale* | The fusion of visible color (RGB) and near | infrared (NIR) images takes multispectral advantage of | colors from RGB image and details from NIR image | . Unlike RGB images, NIR images are robust to atmospheric

摘要： and then perform feature* fusion. However, the* | -train the feature* extraction network from scratch* | resource demands. In this letter, a Feature-selected | Pyramid Network(FsPN) is proposed to address this issue | selection module selects useful information from the

摘要： feature pyramid for finer detail extraction and a cross | -modal dynamic information selection* framework to* | three main modules: a dual ResNet50-based feature | dynamic information selection* module that evaluates the* | Alzheimer's disease (AD) is a common and

摘要： propose a context-aware feature* selection network (CAFS* | network (CANet), a novel feature* selection module (FSM* | ), and a lightweight feature-fusing decoder. The CANet | defective features of image blocks from the feature | pyramid of image blocks based on the classification

摘要： network, Filter-Pruned Bi-directional Feature Pyramid | lightweight detection method based on edge feature* aware and* | feature, we introduce an Edge Feature-Aware (EFA | module. Furthermore, a lightweight feature* fusion* | quantization algorithm based on a bit-width selection

摘要： distance is used for feature* selection on the* | feature redundancy and class imbalance problem in the | feature subsets that can effectively approximate the | original feature* space. Secondly, the loss function, EQL* | , a pyramid* depthwise separable convolution model is*

摘要： feature selection process, which includes the | feature selection. This dynamic approach enhances the | methods rely heavily on predefined feature sets, which | , resulting in a feature-reduced dataset used for training | , specifically the Novel Pyramid* Depth-wise Separable*

摘要：, which use concise head feature* pooling pyramid and* | adaptive feature* selection gate is introduced to reduce* | cost and robust feature* coupling mechanism. However* | In recent years, visual-inertial odometry (VIO | ) has been widely applied in localization, navigation