NSTL国家科技图书文献中心

1. Binarized Simplicial Convolutional Neural Networks NSTL国家科技图书文献中心

Yan, Yi | Kuruoglu, Ercan Engi... - 《Neural Networks》 - 2025,183 - Article 106928～Article 106928 - 共11页

摘要： triangles. Simplicial Convolutional* Neural Networks (SCNN* | Graph Neural* Networks have the limitation of* | architecture named Binarized Simplicial Convolutional Neural | efficiency. In this paper, a novel neural* network* | Networks (Bi-SCNN) is proposed based on the combination

关键词： Graph learning | Graph Neural Networks | Simplicial complex | Convolutional Neural Networks | Binarization

2. TPC track denoising and recognition using convolutional neural networks NSTL国家科技图书文献中心

Gajdos, Matej | da Luz, Hugo Natal ... - 《Computer physics communications》 - 2025,312 - 共9页

摘要： the potential of convolutional* neural networks in* | The capability of convolutional neural | networks to remove spurious signals caused by electronic | of the neural* network is described and its* | resulting from the neural* network and conventional*

关键词： Denoising | Machine learning | Convolutional neural networks | Time projection chambers

3. Self-distillation enhanced adaptive pruning of convolutional neural networks NSTL国家科技图书文献中心

Diao, Huabin | Li, Gongyan ... - 《Pattern Recognition》 - 2025,157 - 共11页

摘要：Convolutional neural networks (CNNs) suffer | from issues of large parameter size and high | computational complexity. To address this, we propose an | adaptive pruning algorithm based on self-distillation | . The algorithm introduces a trainable parameter for

关键词： Convolutional neural networks | Self-distillation | Adaptive pruning

4. Toward Efficient Convolutional Neural Networks With Structured Ternary Patterns NSTL国家科技图书文献中心

Christos Kyrkou - 《IEEE transactions on neural networks and learning systems》 - 2025,36(3) - 5810～5817 - 共8页

摘要： for training. Convolutional neural networks | presents work toward utilizing static convolutional | custom STeP-based networks* that provide good trade-offs* | High-efficiency deep learning (DL) models are | necessary not only to facilitate their use in devices with

关键词： Kernel | Training | Quantization (signal) | Filters | Neural networks | Convolutional neural networks | Convolutional codes

5. Exploiting Gaussian distribution in channel pruning for convolutional neural networks NSTL国家科技图书文献中心

Yuzhou Liu | Bo Liu ... - 《The Journal of Supercomputing》 - 2025,81(5) - 688.1～688.25 - 共25页

摘要： simplifying deep convolutional* neural networks (CNNs). The* | Channel pruning, as an efficient structured | pruning method, is widely used for accelerating and | core of this method lies in selecting a factor to | evaluate the importance of channels and pruning

关键词： Convolutional neural networks | Channel pruning | Gaussian distribution | Regularization

6. STP-CNN: Selection of transfer parameters in convolutional neural networks NSTL国家科技图书文献中心

Otmane Mallouk | Nour-Eddine Joudar ... - 《Expert systems》 - 2025,42(2) - e13728.1～e13728.18 - 共18页

摘要： transfer learning in convolutional* neural networks named* | control exactly which parameters, in each convolutional | Nowadays, transfer learning has shown | promising results in many applications. However, most deep | transfer learning methods such as parameter sharing and

关键词： convolutional neural networks | deep learning | deep transfer learning | 1-regularization | proximal algorithms | transfer learning

7. Boosting Convolutional Neural Networks With Middle Spectrum Grouped Convolution NSTL国家科技图书文献中心

Zhuo Su | Jiehua Zhang ... - 《IEEE transactions on neural networks and learning systems》 - 2025,36(2) - 3436～3449 - 共14页

摘要： efficient deep convolutional* neural networks (DCNNs) with* | This article proposes a novel module called | middle spectrum grouped convolution (MSGC) for | the mechanism of grouped convolution. It explores | the broad middle spectrum area between channel

关键词： Convolution | Computational efficiency | Topology | Convolutional neural networks | Computational modeling | Tensors | Task analysis

8. CONVOLUTIONAL NEURAL NETWORKS FOR IDENTIFYING PAPILLARY THYROID CANCER HISTOPATHOLOGICAL IMAGE NSTL国家科技图书文献中心

NABILA HUSNA SHABRIN... | DADANG GUNAWAN ... - 《International journal of innovative computing,information and control》 - 2025,21(2) - 565～576 - 共12页

摘要： performance, with one of them being Convolutional Neural | Networks (CNN). However, there is a noticeable gap in | Artificial intelligence advancements have | significantly sped up the development of specialized | algorithms for diagnosing Papillary Thyroid Cancer from

关键词： Artificial intelligence | Convolutional neural networks | Histopathological image | Patch size | Papillary thyroid cancer | Transfer learning

9. A concept-aware explainability method for convolutional neural networks NSTL国家科技图书文献中心

Mustafa Kagan Gurkan | Nafiz Arica ... - 《Machine Vision and Applications》 - 2025,36(2) - 33.1～33.17 - 共17页

摘要：Although Convolutional* Neural Networks (CNN* | ) outperform the classical models in a wide range of Machine | Vision applications, their restricted interpretability | and their lack of comprehensibility in reasoning | , generate many problems such as security, reliability, and

关键词： Convolutional neural networks | Concept-based explanation | Filter-concept association | Model comparison via explanations

10. Applying convolutional neural networks for mustard variety recognition NSTL国家科技图书文献中心

Slebioda, Laura | Zawieja, Bogna - 《Euphytica》 - 2025,221(2) - 共11页

摘要： a Convolutional* Neural Network (CNN) model to* | included convolutional* layers, batch normalization* | The aim of this study was to develop and apply | recognize and classify white mustard (Sinapis Alba L | .) varieties, addressing the complex task of discriminating

摘要： triangles. Simplicial Convolutional* Neural Networks (SCNN* | Graph Neural* Networks have the limitation of* | architecture named Binarized Simplicial Convolutional Neural | efficiency. In this paper, a novel neural* network* | Networks (Bi-SCNN) is proposed based on the combination

关键词： Graph learning | Graph Neural Networks | Simplicial complex | Convolutional Neural Networks | Binarization

摘要： the potential of convolutional* neural networks in* | The capability of convolutional neural | networks to remove spurious signals caused by electronic | of the neural* network is described and its* | resulting from the neural* network and conventional*

关键词： Denoising | Machine learning | Convolutional neural networks | Time projection chambers

摘要：Convolutional neural networks (CNNs) suffer | from issues of large parameter size and high | computational complexity. To address this, we propose an | adaptive pruning algorithm based on self-distillation | . The algorithm introduces a trainable parameter for

关键词： Convolutional neural networks | Self-distillation | Adaptive pruning

摘要： for training. Convolutional neural networks | presents work toward utilizing static convolutional | custom STeP-based networks* that provide good trade-offs* | High-efficiency deep learning (DL) models are | necessary not only to facilitate their use in devices with

关键词： Kernel | Training | Quantization (signal) | Filters | Neural networks | Convolutional neural networks | Convolutional codes

摘要： simplifying deep convolutional* neural networks (CNNs). The* | Channel pruning, as an efficient structured | pruning method, is widely used for accelerating and | core of this method lies in selecting a factor to | evaluate the importance of channels and pruning

关键词： Convolutional neural networks | Channel pruning | Gaussian distribution | Regularization

摘要： transfer learning in convolutional* neural networks named* | control exactly which parameters, in each convolutional | Nowadays, transfer learning has shown | promising results in many applications. However, most deep | transfer learning methods such as parameter sharing and

关键词： convolutional neural networks | deep learning | deep transfer learning | 1-regularization | proximal algorithms | transfer learning

摘要： efficient deep convolutional* neural networks (DCNNs) with* | This article proposes a novel module called | middle spectrum grouped convolution (MSGC) for | the mechanism of grouped convolution. It explores | the broad middle spectrum area between channel

关键词： Convolution | Computational efficiency | Topology | Convolutional neural networks | Computational modeling | Tensors | Task analysis

摘要： performance, with one of them being Convolutional Neural | Networks (CNN). However, there is a noticeable gap in | Artificial intelligence advancements have | significantly sped up the development of specialized | algorithms for diagnosing Papillary Thyroid Cancer from

关键词： Artificial intelligence | Convolutional neural networks | Histopathological image | Patch size | Papillary thyroid cancer | Transfer learning

摘要：Although Convolutional* Neural Networks (CNN* | ) outperform the classical models in a wide range of Machine | Vision applications, their restricted interpretability | and their lack of comprehensibility in reasoning | , generate many problems such as security, reliability, and

关键词： Convolutional neural networks | Concept-based explanation | Filter-concept association | Model comparison via explanations

摘要： a Convolutional* Neural Network (CNN) model to* | included convolutional* layers, batch normalization* | The aim of this study was to develop and apply | recognize and classify white mustard (Sinapis Alba L | .) varieties, addressing the complex task of discriminating

关键词： Agricultural algorithms | Classification | Convolutional neural networks | Mustard | Variety recognition

4008-161-200 800-990-8900

国家科技图书文献中心

摘要： triangles. Simplicial Convolutional Neural Networks (SCNN | Graph Neural Networks have the limitation of | architecture named Binarized Simplicial Convolutional Neural | efficiency. In this paper, a novel neural network | Networks (Bi-SCNN) is proposed based on the combination

关键词： Graph learning | Graph Neural Networks | Simplicial complex | Convolutional Neural Networks | Binarization

摘要： the potential of convolutional neural networks in | The capability of convolutional neural | networks to remove spurious signals caused by electronic | of the neural network is described and its | resulting from the neural network and conventional

关键词： Denoising | Machine learning | Convolutional neural networks | Time projection chambers

摘要：Convolutional neural networks (CNNs) suffer | from issues of large parameter size and high | computational complexity. To address this, we propose an | adaptive pruning algorithm based on self-distillation | . The algorithm introduces a trainable parameter for

关键词： Convolutional neural networks | Self-distillation | Adaptive pruning

摘要： for training. Convolutional neural networks | presents work toward utilizing static convolutional | custom STeP-based networks that provide good trade-offs | High-efficiency deep learning (DL) models are | necessary not only to facilitate their use in devices with

关键词： Kernel | Training | Quantization (signal) | Filters | Neural networks | Convolutional neural networks | Convolutional codes

摘要： simplifying deep convolutional neural networks (CNNs). The | Channel pruning, as an efficient structured | pruning method, is widely used for accelerating and | core of this method lies in selecting a factor to | evaluate the importance of channels and pruning

关键词： Convolutional neural networks | Channel pruning | Gaussian distribution | Regularization

摘要： transfer learning in convolutional neural networks named | control exactly which parameters, in each convolutional | Nowadays, transfer learning has shown | promising results in many applications. However, most deep | transfer learning methods such as parameter sharing and

关键词： convolutional neural networks | deep learning | deep transfer learning | 1-regularization | proximal algorithms | transfer learning

摘要： efficient deep convolutional neural networks (DCNNs) with | This article proposes a novel module called | middle spectrum grouped convolution (MSGC) for | the mechanism of grouped convolution. It explores | the broad middle spectrum area between channel

关键词： Convolution | Computational efficiency | Topology | Convolutional neural networks | Computational modeling | Tensors | Task analysis

摘要： performance, with one of them being Convolutional Neural | Networks (CNN). However, there is a noticeable gap in | Artificial intelligence advancements have | significantly sped up the development of specialized | algorithms for diagnosing Papillary Thyroid Cancer from

关键词： Artificial intelligence | Convolutional neural networks | Histopathological image | Patch size | Papillary thyroid cancer | Transfer learning

摘要：Although Convolutional Neural Networks (CNN | ) outperform the classical models in a wide range of Machine | Vision applications, their restricted interpretability | and their lack of comprehensibility in reasoning | , generate many problems such as security, reliability, and

关键词： Convolutional neural networks | Concept-based explanation | Filter-concept association | Model comparison via explanations

摘要： a Convolutional Neural Network (CNN) model to | included convolutional layers, batch normalization | The aim of this study was to develop and apply | recognize and classify white mustard (Sinapis Alba L | .) varieties, addressing the complex task of discriminating

关键词： Agricultural algorithms | Classification | Convolutional neural networks | Mustard | Variety recognition

4008-161-200 800-990-8900

国家科技图书文献中心

摘要： triangles. Simplicial Convolutional* Neural Networks (SCNN* | Graph Neural* Networks have the limitation of* | architecture named Binarized Simplicial Convolutional Neural | efficiency. In this paper, a novel neural* network* | Networks (Bi-SCNN) is proposed based on the combination

摘要： the potential of convolutional* neural networks in* | The capability of convolutional neural | networks to remove spurious signals caused by electronic | of the neural* network is described and its* | resulting from the neural* network and conventional*

摘要： for training. Convolutional neural networks | presents work toward utilizing static convolutional | custom STeP-based networks* that provide good trade-offs* | High-efficiency deep learning (DL) models are | necessary not only to facilitate their use in devices with

摘要： simplifying deep convolutional* neural networks (CNNs). The* | Channel pruning, as an efficient structured | pruning method, is widely used for accelerating and | core of this method lies in selecting a factor to | evaluate the importance of channels and pruning

摘要： transfer learning in convolutional* neural networks named* | control exactly which parameters, in each convolutional | Nowadays, transfer learning has shown | promising results in many applications. However, most deep | transfer learning methods such as parameter sharing and

摘要： efficient deep convolutional* neural networks (DCNNs) with* | This article proposes a novel module called | middle spectrum grouped convolution (MSGC) for | the mechanism of grouped convolution. It explores | the broad middle spectrum area between channel

摘要：Although Convolutional* Neural Networks (CNN* | ) outperform the classical models in a wide range of Machine | Vision applications, their restricted interpretability | and their lack of comprehensibility in reasoning | , generate many problems such as security, reliability, and

摘要： a Convolutional* Neural Network (CNN) model to* | included convolutional* layers, batch normalization* | The aim of this study was to develop and apply | recognize and classify white mustard (Sinapis Alba L | .) varieties, addressing the complex task of discriminating