检索结果-内蒙古大学图书馆

Chinese Control and Decision Conference, CCDC

作者： Xiaoliang Lei Xiaosheng Yu Maocheng Bai Chengdong Wu College of Information Science and Engineering Northeastern University Shenyang China Faculty of Robot Science and Engineer Northeastern University Shenyang China

ISBN: (数字)9798350387780

ISBN: (纸本)9798350387797

Brain MRI synthesis technology addresses the challenge of missing MRI modalities in the clinical domain. We strongly emphasize harnessing the full potential of multi-modal MRI data and the spatial correlations within brain structures, so we proposed a method for synthesizing brain MRI images. This method distills latent information from available MRI modalities, providing guidance for synthesizing the missing MRI modality, thereby transcending the constraints of spatial structural relevance within the task of 3D brain MRI image synthesizing. Our experiments demonstrate that our method can generate high-quality 3D brain MRI images.

关键词： Three-dimensional displays Correlation Magnetic resonance imaging Diffusion processes Transforms Aerospace electronics Diffusion models

来源：评论

学校读者我要写书评

暂无评论

A Face Super-Resolution Reconstruction Algorithm Based on Residual Estimation

A Face Super-Resolution Reconstruction Algorithm Based on Re...

引用

IEEE International Conference on Cyber Technology in Automation, Control, and Intelligent Systems

作者： Xiaosheng Yu Guanglai Liu Zi Teng Xiaoqiang Li Jubo Chen Faculty of Robot Science and Engineering Northeastern University Shenyang Liaoning China College of Information Science and Engineering Northeastern University Shenyang Liaoning China

ISBN: (数字)9798331506056

ISBN: (纸本)9798331506063

Face super resolution can greatly improve the performance of various facial analysis applications, such as face recognition and facial expression analysis. This paper introduces a facial super-resolution network model that utilizes residual estimation for enhancing image quality. The model engages a two-stage reconstruction process to generate super resolution face images, which effectively simplifies the learning complexities of the network. Moreover, the model incorporates prior facial information into the loss function to mitigate the influence of the background on the facial region within the image. This incorporation facilitates a more accurate reconstruction of super resolution facial images. The robustness and effectiveness of this facial super-resolution network based on residual estimation are evaluated through both quantitative and qualitative assessments, employing some classical facial image datasets.

关键词： Image quality Face recognition Image edge detection Superresolution Estimation Reconstruction algorithms Robustness Intelligent systems Image reconstruction Faces

来源：评论

学校读者我要写书评

暂无评论

Utilizing Large Language Models Enhanced by Chain-of-Thought for the Diagnosis of Typical Medical Cases 10th

Utilizing Large Language Models Enhanced by Chain-of-Though...

引用

10th China Health Information Processing Conference, CHIP 2024

作者： Liu, Jiqiang Liu, Chenyang Faculty of Robot Science and Engineering Northeastern University Shenyang China School of Computer Science and Engineering Northeastern University Shenyang China

ISBN: (纸本)9789819642977

In the past two years, the large language model has set off a new wave of research in the field of natural language processing, showing the ability of general-purpose artificial intelligence, which has been widely concerned by the industry. With the rapid development of medical vertical large model, its potential in clinical application has been paid more and more attention. However, there are gaps in the diagnosis of typical medical records. In order to continuously improve the application effect of medical large model in actual clinical scenarios and accelerate the implementation of medical large model, this paper describes the specific situation of our participation in the 10th China Conference on Health Information Processing (CHIP 2024). We used the improved LoRA method for fine tuning, and then used the Chain-of-Thought method for post-processing in the reasoning stage to improve performance. The experimental results show that the F1 score of our method in the final second round of evaluation reaches 0.9356, ranking second, which effectively verifies the generalization and robustness of our method. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Diagnosis

来源：评论

学校读者我要写书评

暂无评论

Divided Block Multiscale Convolutional Network for Micro-expression Recognition

Divided Block Multiscale Convolutional Network for Micro-exp...

引用

Cyber-Energy Systems and Intelligent Energy (ICCSIE), International Conference on

作者： Quan Zhou Shiyu Liu Yiheng Wang Junyi Wang Faculty of Robot Science and Engineering Northeastern University Shenyang China Faculty of Robot Science and Engineering and Foshan Graduate School of Innovation Northeastern University Shenyang China

Micro-expression (ME) is a subtle change in the face, which can be used to judge human subjective feelings. It has broad application prospects in medical diagnosis and business negotiation. However, due to the complexity of ME muscle movements and the lack of ME trainable data, the research of micro-expression recognition (MER) still faces a series of challenges. In this paper, we propose a new divided block multiscale convolution network (DBMNet), which could learn from four different optical flow (OF) feature images obtained between the onset and apex frames of ME samples. Through the proposed block-divided multiscale convolution module (BMCM), more detailed and useful multiscale advanced features of ME could be effectively extracted. In order to better address the problem of class imbalance on the ME dataset, this paper uses the weighted cross entropy (CE) loss function, which could obviously alleviate the impact of class imbalance. Finally, 5-class experiment is conducted on the composite dataset to show that the proposed method has superior performance and is comparable to those of the most advanced methods.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Research on Bank Card Number Recognition Based on ASFF+YOLOv7 and Multi-Scale Feature Line Fusion 13

A Research on Bank Card Number Recognition Based on ASFF+YOL...

引用

13th IEEE International Conference on CYBER Technology in Automation, Control, and Intelligent Systems, CYBER 2023

作者： Cheng, Jiajun Wu, Yaoyan Gao, Weitao Fu, Jun Northeastern University Faculty of Robot Science and Engineering Liaoning Shenyang China School of Electrical and Control Engineering State Key Laboratory of Dynamic Testing Technology North University of China Shanxi Taiyuan China

ISBN: (纸本)9798350315196

Electronic payments have become the primary mode of payment today, and bank card recognition is widely used in industries such as mobile, mobile banking, and third-party payments. To address issues such as low fault tolerance due to factors such as complex backgrounds, significant perspective distortion that is difficult to recover, and high illumination variation, we propose a Bank Card Number Recognition algorithm based on ASFF + YOLOv7 and Multi-Scale Feature Line Fusion. The algorithm comprises two parts: the front-end dataset generation stage introduces an efficient semi-automatic traditional bank card dataset generation algorithm, which reduces the manual annotation cost while ensuring relative accuracy and significantly improves work efficiency. In the back-end recognition stage, we have improved the traditional deep learning model for the long-tail digit string dataset, achieving a precision of 99.19% and a recall rate of 98.65%. © 2023 IEEE.

关键词： Fault tolerance

来源：评论

学校读者我要写书评

暂无评论

HierNet: Hierarchical Transformer U-Shape Network for RGB-D Salient Object Detection

HierNet: Hierarchical Transformer U-Shape Network for RGB-D ...

引用

Chinese Control and Decision Conference, CCDC

作者： Pengfei Lv Xiaosheng Yu Junxiang Wang Chengdong Wu Faculty of Robot Science and Engineering Northeastern University Shenyang China

With the popularity of depth sensors, research on RGB-D salient object detection (SOD) is also thriving. However, given the limitations of the external environment and the sensor itself, depth information is often less credible. To meet this challenge, existing models often purify the depth information using complex convolution and pooling operations. This causes a large amount of useful information besides noise to be dropped as well, and multi-modality interaction chances between RGB and depth become less. Also, with the gradual loss of information, the hidden relationship of features between multi-level is thus ignored. To tackle the aforementioned problems, we propose a Hierarchical Transformer U-Shape Network (HierNet) that include three aspects: 1) With a simple structure, a depth calibration module provides faithful depth information with minimal loss of information, providing conditions for cross-modality cross-layer information interaction; 2) With multi-head attention, a set of global view-based transformer encoders are employed to find the potential coherence between RGB and depth modalities. With weight sharing, several transformer encoder sets comprise the hierarchical transformer embedding module to search long-range dependencies cross-level; 3) Considering the complementary features of U-shape network, we use dual-stream U-shape network as our backbone. Extensive fair experiments on four challenging datasets have demonstrated the outstanding performance of the proposed model compared to state-of-the-art models.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Efficient Object Localization for Unseen Object 6D Pose Estimation

Efficient Object Localization for Unseen Object 6D Pose Esti...

引用

Chinese Automation Congress (CAC)

作者： Xinwei Lan Chengdong Wu Xiangyue Zhang Faculty of Robot Science and Engineering Northeastern University Shenyang China

Object localization is utilized as the first step in standard 6D object pose estimation methods to obtain the position information of the objects. However, these object localization methods cannot be directly applied to unseen objects, which is the focus of recent research on 6D object pose estimation. In this paper, an accurate and efficient localization method for unseen object is proposed, based on a template matching strategy. The Hybrid Channel-Spatial Attention Model (HCSAM) is designed to focus on the target object by enhancing the contextual differences between the target object and background. Additionally, The Multi-Scale Integration Transformer (MSIT) module is designed to eliminate noise interference and enhance semantic information in low-dimensional features by integrating multidimensional information. Our method outperforms existing approaches on the complicated occluded dataset LINEMOD, as well as on the challenging generalized pose estimation dataset GenMOP.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Binary Classification is Enough: A Lightweight Strategy for Drug Screening with Small Datasets

Binary Classification is Enough: A Lightweight Strategy for ...

引用

Health Big Data and Intelligent Healthcare (ICHIH), International Conference on

作者： Liang Wu Xiaoguang Ma Faculty of Robot Science and Engineering Northeastern University Shenyang China

Drug screening is an extremely costly and time-consuming process, wherein only small datasets are available in practice. We presented a particular method to estimate values of inhibition constant(Ki) or half-maximal inhibition concentration(IC50) of unknown compounds through a lightweight mutual information and logistic regression(MI-LR) united model that only needed to be trained on a small dataset. Biologists could then use this model to determine whether the compounds were initially eligible for screening, increasing efficiency of their work. A data augmentation strategy was used to sort independent samples of training datasets and solved the problem of sample shortage caused by the lightweight model, and transform a prediction task into a simpler binary classification task. In addition, we proposed an effective constraint mechanism to deal with the case when the classification results were contrary to the facts. By accurately predicting the interval of its inhibitory effect, we can improve the efficiency and accuracy of drug screening. Numerous evaluations on the Ki and IC50 dataset demonstrated high reliability of the MI-LR united approach to sort compounds according to a selected set of molecular descriptors.

关键词：

来源：评论

学校读者我要写书评

暂无评论

MMF-Track: Multi-modal Multi-level Fusion for 3D Single Object Tracking

arXiv

引用

arXiv 2023年

作者： Li, Zhiheng Cui, Yubo Lin, Yu Fang, Zheng The Faculty of Robot Science and Engineering Northeastern University Shenyang China

3D single object tracking plays an important role in computer vision and autonomous driving. The mainstream methods mainly rely on point clouds to achieve geometry matching between target template and search area. However, textureless and incomplete point clouds make it difficult for single-modal trackers to distinguish objects with similar structures. To overcome the mentioned limitations of geometry matching, we propose a Multi-modal Multi-level Fusion Tracker (MMF-Track), which exploits the image texture and geometry characteristic of point clouds to track 3D target. Specifically, we first propose a Space Alignment Module (SAM) to align RGB images with point clouds in 3D space, which is the prerequisite for constructing inter-modal associations. After that, in feature interaction level, we present a Feature Interaction Module (FIM) based on dual-stream structure, which enhances intra-modal features in parallel and constructs inter-modal semantic associations. Meanwhile, in order to refine each modal feature, we propose a Coarse-to-Fine Interaction Module (CFIM) to realize the hierarchical feature interaction at different scales. Finally, in similarity fusion level, we introduce a Similarity Fusion Module (SFM) to aggregate geometry and texture similarity from the target. Extensive experiments show that our method achieves competitive performance on KITTI and NuScenes datasets. The code will be opened soon in https://***/LeoZhiheng/***. Copyright © 2023, The Authors. All rights reserved.

关键词： Data fusion

来源：评论

学校读者我要写书评

暂无评论

A Novel Generalized EEG Channel Selection Method Using Pearson Correlation Coefficient*

A Novel Generalized EEG Channel Selection Method Using Pears...

引用

IEEE International Conference on robotics and Biomimetics

作者： Dongxu Liu Qichuan Ding Maiwei Wen Chenyu Tong Faculty of Robot Science and Engineering Northeastern University Shenyang China

Electroencephalography (EEG), as a non-invasive and convenient method for implementing Brain-Computer Interface (BCI), has been widely used in clinical and research fields. EEG data often requires the acquisition of dozens or even hundreds of channels. Channel selection can reduce irrelevant and redundant channels, improve computational efficiency, and enhance the quality of EEG signals. This study introduces a filter method for channel selection based on Pearson correlation coefficient (PCC) with the candidate channel and employs topographic maps of EEG channel scores, derived from data collected across all subjects, to visualize the spatial distribution of channels selected by different methods. In addition, a generalized channel selection algorithm is proposed to determine consistent channels across all subjects in the experimental group. The effectiveness of the proposed method was evaluated on two steady-state visual evoked potential (SSVEP) datasets, and the results indicated that this method exhibits superior performance compared to both the all-channel method and other channel selection methods. And the application of the generalized channel algorithm has further improved the classification performance. This study uses selected generalized channels applied to new subjects with low BCI performance, yielding a significant improvement. The selected channels have a wide range of applicability, helping to simplify EEG acquisition and improve EEG data quality.

关键词：

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：