检索结果-内蒙古大学图书馆

2023 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2023

作者： Zhu, Qinglian Lin, Xiaoli Zhang, Xiaolong Wuhan University of Science and Technology Hubei Key Laboratory of Intelligent Information Processing and Real-Time Industrial System School of Computer Science and Technology Wuhan China

ISBN: (纸本)9798350337488

Most current representations of protein pockets are the atom-pair graph, which ignore the global structural information of amino acids. Therefore, we propose a new molecular generation model, which uses the hypergraph to represent protein pocket structure, and combines the structural features obtained by atom-pair graph representation. These two levels of graphs are more capable of representing the complex structural information of protein pockets. Then, the graphs of the two levels of the protein pockets are input into the improved network model, which is named Hypergraph Graph Attention Fusion (HGAF), to obtain the embedding representation of the protein pockets, which is used as a condition to constrain the molecule generation. The molecules sampled by HGAF are subjected to quality assessment and docking targeting validation. Experimental results show that the molecules generated by proposed method can achieve better results in both of these assessment approaches. © 2023 IEEE.

关键词： 3D Protein Pockets Drug Design Graph Attention Networks

来源：评论

学校读者我要写书评

暂无评论

How to Train Your Neural Network for Molecular Structure Generation from Mass Spectra?

How to Train Your Neural Network for Molecular Structure Gen...

引用

2024 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2024

作者： Zhao, Kai Liu, Yanmin Dian, Longyang Sun, Shiwei Cui, Xuefeng Shandong University School of Computer Science and Technology Qingdao China Shandong University State Key Laboratory of Microbial Technology Qingdao China Chinese Academy of Sciences Key Lab of Intelligent Information Processing Institute of Computing Technology Beijing China

ISBN: (纸本)9798350386226

Mass spectrometry serves as a pivotal tool for the analysis of small molecules through an examination of their mass-to-charge ratios. Recent advancements in deep learning have markedly enhanced the analysis of mass spectrometric data, facilitating the prediction of novel small molecule structures without the necessity of extensive databases. Nonetheless, the paucity of annotated datasets impedes the efficacious training of molecular generation models predicated on MS2 spectra. To mitigate this limitation, we introduce ctMSNovelist, an avant-garde method that amalgamates pre-training, fine-tuning, and co-training techniques to construct a more precise model for the generation of molecular structures from tandem mass spectrometry data. This novel approach augments both the training regimen and the predictive accuracy of the MSNovelist model, thereby surmounting the obstacle of limited data. The methodology commences with the pretraining of a Variational Autoencoder (VAE) to generate molecular fingerprints derived from SMILES strings. Subsequently, it undergoes fine-tuning to emulate noisy fingerprints originating from mass spectrometry (MS) data. Concurrently, MSNovelist is co-trained utilizing these simulated fingerprints, inclusive of the highly noisy variants produced in the early stages of VAE training. The incorporation of a substantial volume of noisy data serves to enhance model accuracy and avert overfitting. We evaluated ctMSNovelist using the GNPS dataset and attained a SMILES prediction accuracy of 48.8%, representing a 4.1% enhancement over MSNovelist. It is pertinent to note that the sole distinction between ctMSNovelist and MSNovelist in this experiment was the training process. The code and models are publicly available at https://***/xfcui/ctMSNovelist. © 2024 IEEE.

关键词： Neural networks

来源：评论

学校读者我要写书评

暂无评论

Anti-3CLpro Molecular Design Based on the Model Constrained by Specific DTIs

Anti-3CLpro Molecular Design Based on the Model Constrained ...

引用

2023 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2023

作者： Lin, Xiaoli Liu, Xuan Zhang, Xiaolong School of Computer Science and Technology Wuhan University of Science and Technology Hubei Key Laboratory of Intelligent Information Processing and Real-Time Industrial System Wuhan China

ISBN: (纸本)9798350337488

computer-aided drug design and artificial intelligence-driven drug design have accelerated drug discovery. However, how to design effective drugs that have strong interaction ability with target proteins to further improve the efficacy of drugs in treating diseases remains a key issue. This paper proposes a new target-specific drug generative model 3CLpro2mol to generate new drug molecules, which uses features of drug-target interactions (DTIs) to constrain the correlation between the drug and the target protein. To obtain as many drug-target interaction features as possible from a small amount of data, a small molecule extraction strategy is proposed to ensure the diversity of small molecules in the training samples. To improve the efficiency and accuracy of the generative model, a TOP K sampling strategy is used to generate tokens, which can improve the rationality and diversity of the generated molecules. The experimental results show that the proposed model has the potential to generate small molecules that interact better with the target protein. © 2023 IEEE.

关键词： COVID-19 DTIs generative model molecular design molecular docking

来源：评论

学校读者我要写书评

暂无评论

A Convolutional Knowledge Tracing Method based on Graph Embedding 4

A Convolutional Knowledge Tracing Method based on Graph Embe...

引用

4th International Symposium on computer Engineering and Intelligent Communications, ISCEIC 2023

作者： Huang, Guimin Li, Ying Guo, Qingkai Deng, Nanxiao Lin, Jianxing Guilin University of Electronic Technology School of Computer Science and Information Security Guangxi Key Laboratory of Image and Graphic Intelligent Processing Guilin China

ISBN: (纸本)9798350306323

With online education rapidly developing, it is a significant issue to evaluate students' grasp of knowledge more accurately. Knowledge tracing models are good at it, in which convolutional knowledge tracing (CKT) based on convolutional neural networks(CNN) gets better performance. However, we found that the CKT model ignores the relation between the knowledge points, which we believe has an important impact on the correctness of students' answers. Therefore, we put forward a graph embedding based convolutional knowledge tracing model (CKT-GE), which feeds the relation between the knowledge points with other inputs into model for improving the effectiveness of knowledge tracing. Our model outperforms existing models on two classic datasets, and improves the performance over the original CKT model, allowing for more accurate assessment of student performance. © 2023 IEEE.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

More Accurate Constraints for Self-Supervised Learning in Remote Sensing Images based Object Detection

引用

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 2025年 18卷 12303-12314页

作者： Zheng, Shangdong Wu, Zebin Xu, Yang Liu, Qian Wei, Zhihui Nanjing University of Posts and Telecommunications School of Computer Science Nanjing210023 China Nanjing University of Science and Technology School of Computer Science and Engineering Nanjing210094 China Nanjing University of Science and Technology Jiangsu Key Laboratory of Spectral Imaging & Intelligent Sense China Anhui Agricultural University School of Information and Artificial Intelligence Hefei230036 China

Self-supervised learning (SSL) automatically generates internal labels by exploring the potential auxiliary task of the network itself, and trains the same model through these annotations to learn the latent representation of the data, which greatly improves the accuracy of object detection in remote sensing images (RSIs). However, most existing methods suffer from guaranteeing the quality of the generated SSL pseudo annotations, and the constructed auxiliary tasks are not detectionoriented, which is difficult to enhance the feature representations that are beneficial for object detection. In this paper, we focus on generating more accurate constraints by excavating the intercorrelation between fully and weakly-supervised learning to improve the performance of object detection in RSIs. Initially, weakly-supervised learning (WSL) assigns the pseudo instancelevel annotations for the high-scoring positive bags to model the detector, which can be regarded as a weakened version of the region proposal network (RPN). Fortunately, RPN can be constrained by the ground truth of bounding boxes in fullysupervised learning (FSL), and the high-quality supervisions it provides are unavailable in any WSL methods. Moreover, we construct a proposal generation module (PGM), which further filters the unreliable bounding boxes, predicted by RPN, to supplement high-quality constraints into the ground truth and SS generated candidate boxes to supervise the optimization of WSL branch. By constructing an interactive learning paradigm of WSL and FSL, the former has more accurate constraints to learn an efficient auxiliary task, while the latter enjoys a richer representation form of data provided by WSL, which is undoubtedly a win-win process. Finally, we cascade the losses of WSL and FSL to further explore the intrinsic correlation between them by sharing the same feature extraction network. Experimental comparisons on DOTA and DIOR datasets demonstrate that our method achieves superior per

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Improving Speaker Verification Back-End with Graph Neural Networks

引用

Journal of Shanghai Jiaotong University (Science) 2025年 1-9页

作者： Chen, Jinfeng Fang, Zhihua He, Liang School of Computer Science and Technology Xinjiang University Urumqi830017 China Xinjiang Key Laboratory of Signal Detection and Processing Urumqi830017 China School of Intelligence Science and Technology Xinjiang University Urumqi830017 China Department of Electronic Engineering Beijing National Research Center for Information Science and Technology Tsinghua University Beijing100084 China

Currently, research on speaker verification tasks is primarily concentrated on enhancing deep speaker models to extract high-quality speaker embeddings. Nevertheless, this speaker embeddings can be regarded as potential graph structures. Therefore, this paper introduces a graph neural network-based speaker verification back-end approach, treating the speaker embeddings derived from the front-end as graph structures and leveraging graph neural networks to uncover the relationships among embeddings, thereby achieving superior-quality speaker embeddings. In addition, we propose a group updating method to solve the problem of memory overflow when the number of nodes is excessive. Extensive experiments and ablation studies were carried out on the VoxCeleb dataset, and the experimental outcomes validate the effectiveness of our proposed graph neural network speaker back-end in significantly boosting the performance of speaker verification systems. (Figure presented.) © Shanghai Jiao Tong University 2025.

关键词： Graph embeddings

来源：评论

学校读者我要写书评

暂无评论

A Deep Knowledge Tracking Model Based on Forgetting Features and Self-Attention Mechanism 4

A Deep Knowledge Tracking Model Based on Forgetting Features...

引用

4th International Symposium on computer Engineering and Intelligent Communications, ISCEIC 2023

作者： Huang, Guimin Wei, Xiong Guo, Qingkai Deng, Nanxiao Lin, Jianxing Guilin University of Electronic Technology School of Computer Science and Information Security Guangxi Key Laboratory of Image and Graphic Intelligent Processing Guilin China

ISBN: (纸本)9798350306323

The integration of large education and artificial intelligence technologies is gradually deepening, and how to provide personalized user profiling services for learners is an important research problem. In response to the long-term dependency problem and lack of learning features of today's deep knowledge tracking models, a deep knowledge tracking model (FSA-MuLSTM) based on forgetting features and the self-attentiveness mechanism is proposed. Firstly, the model combines the learner's historical learning interaction tuple sequences with learning times and memory residuals to obtain the input vector based on the LSTM;secondly, the correlation weights between tuples are captured using the self-attention mechanism;finally, the deep information between tuple sequences is captured using the stacking of multilayer LSTMs. The probability of correctly answering the current question is derived from the hidden vectors and the self-attentive weights of the output of the multilayer LSTM. Two public datasets are used for the experiments, and the combined analysis outperforms existing knowledge-tracking models in terms of performance and improves the accuracy of prediction. © 2023 IEEE.

关键词： Multilayers

来源：评论

学校读者我要写书评

暂无评论

Anomalous Sound Detection Using Time-Frequency Feature and Mixbatch

引用

Journal of Shanghai Jiaotong University (Science) 2025年 1-8页

作者： Huang, Shun Zhang, Yunxiang Fang, Zhihua Tang, Minrui Xu, Ruifeng He, Liang School of Computer Science and Technology Xinjiang University Urumqi830017 China Xinjiang Key Laboratory of Signal Detection and Processing Urumqi830017 China School of Intelligence Science and Technology Xinjiang University Urumqi830017 China Department of Electronic Engineering Beijing National Research Center for Information Science and Technology Tsinghua University Beijing100084 China

The sound emitted by machines under abnormal working conditions exhibits various frequency patterns. Currently, the most advanced anomalous sound detection (ASD) approach is to apply a multi-head self-attention mechanism to the Log-Mel spectrogram for automatic frequency pattern analysis. However, the Log-Mel spectrogram may filter out high-frequency components of abnormal sounds;thus the use of self-attention mechanisms on Log-Mel spectrogram seems to have certain limitations. In this paper, we construct a simple convolutional neural network to extract comprehensive frequency features from raw audio to complement spectral-temporal information fusion. The parameters of this neural network are continuously updated during the training process to extract better frequency features for downstream classification neural networks in ASD. Additionally, a method for data augmentation in the batch dimension was developed to help the classification model learn both types of comprehensive features simultaneously. The proposed method achieved an AUC of 94.13% and a pAUC of 89.09% on the DCASE 2020 Challenge Task 2 dataset. Even when using only the features proposed in this paper, an AUC of 84.62% was achieved. (Figure presented.). © Shanghai Jiao Tong University 2025.

关键词： Spectrographs

来源：评论

学校读者我要写书评

暂无评论

DensityLayout: Density-Conditioned Layout GAN for Visual-Textual Presentation Designs 12th

DensityLayout: Density-Conditioned Layout GAN for Visual-Te...

引用

12th International Conference on Image and Graphics, ICIG 2023

作者： Hsu, HsiaoYuan He, Xiangteng Peng, Yuxin Wangxuan Institute of Computer Technology Peking University Beijing China National Key Laboratory for Multimedia Information Processing Peking University Beijing China

ISBN: (纸本)9783031463075

Generating layouts for visual-textual presentation designs aims at arranging elements such as logo, text, and underlay on the given images, which is the key to automating poster designs. It is challenging since the compositions of images, the spatial patterns of layout elements, and their cross-relationships need to be simultaneously considered. Existing works focus on the cross-relationships and either (1) suffer from the instability of adopting off-the-shelf saliency maps as prior knowledge or (2) require the semantic content of each element. To this end, this paper presents an efficient density paradigm that requires neither off-the-shelf models nor additional training data other than image-layout pairs. Under this paradigm, a three-stage approach using GAN is proposed, entitled DensityLayout. First, a density mapping network weakly supervised by the custom consistency loss will translate given images to spatial distributions of elements. Second, a multi-scale strategy is proposed to enhance understanding of the maps, and a generator conditioned on these visual features will generate preliminary layouts. Finally, a directed graph representation illustrating the inclusion relationships between elements is presented, and a graph convolution network will fine-tune the layouts. The effectiveness of the proposed approach is validated on CGL-Dataset, showing it achieves the best performance by generating visually appealing layouts for visual-textual presentation designs of diverse images. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.

关键词： Directed graphs

来源：评论

学校读者我要写书评

暂无评论

Non-intrusive load monitoring based on a combination of Transformer and CNN 4

Non-intrusive load monitoring based on a combination of Tran...

引用

4th International Conference on Energy, Power and Electrical Engineering, EPEE 2024

作者： Cheng, Yiqi Zhong, Yanru Guilin University of Electronic Technology College of Computer and Information Security Guangxi Guilin China Guilin University of Electronic Technology Guangxi Key Laboratory of Image And Graphic Intelligent Processing Guangxi Guilin China

ISBN: (纸本)9798331518066

Non-intrusive load monitoring (NILM) estimates the energy usage of individual devices by analyzing the total power meter data from a household. To address the issue of decreased monitoring accuracy with an increasing number of electricity meters, this paper proposes a solution for single appliance load monitoring (Tra-CNN). The solution uses the UK-DALE dataset, which contains high and low frequencies electricity consumption data of multiple households in the UK and the electricity consumption data of each household appliance, and uses the active power of the appliance as the input of the algorithm. The first part of the model consists of a convolutional layer, a normalization layer, and a dropout layer, with the primary task of capturing local dependencies in the data. The second part consists of a Transformer, whose self-attention mechanism associates features at different positions in the sequence to achieve modeling of long-term dependencies. The third part introduces a convolutional neural network, which uses a sliding window for convolution operations to extract local features from different positions and further extract features from sequence data. The final decoder consists of a transposed convolutional layer and a normalization layer, designed to revert the learned features to their original dimensions. Through experimental verification, the effectiveness and superiority of Tra-CNN are shown, which provides important ideas and methods for the further development of NILM technology. © 2024 IEEE.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：