检索结果-内蒙古大学图书馆

arXiv 2022年

作者： Fu, Kexue Yuan, Mingzhi Wang, Manning Digital Medical Research Center School of Basic Medical Science Fudan University Shanghai China Shanghai Key Laboratory of Medical Image Computing and Computer Assisted Intervention Shanghai China

Masked language modeling (MLM) has become one of the most successful self-supervised pre-training task. Inspired by its success, Point-BERT, as a pioneer work in point cloud, proposed masked point modeling (MPM) to pre-train point transformer on large scale unanotated dataset. Despite its great performance, we find the inherent difference between language and point cloud tends to cause ambiguous tokenization for point cloud. For point cloud, there doesn’t exist a gold standard for point cloud tokenization. Point-BERT use a discrete Variational AutoEncoder (dVAE) as tokenizer, but it might generate different token ids for semantically-similar patches and generate the same token ids for semantically-dissimilar patches. To tackle above problem, we propose our McP-BERT, a pre-training framework with multi-choice tokens. Specifically, we ease the previous single-choice constraint on patch token ids in Point-BERT, and provide multi-choice token ids for each patch as supervision. Moreover, we utilitze the high-level semantics learned by transformer to further refine our supervision signals. Extensive experiments on point cloud classification, few-shot classification and part segmentation tasks demonstrate the superiority of our method, e.g., the pre-trained transformer achieves 94.1% accuracy on ModelNet40, 84.28% accuracy on the hardest setting of ScanObjectNN and new state-of-the-art performance on few-shot learning. We also demonstrate that our method not only improves the performance of Point-BERT on all downstream tasks, but also incurs almost no extra computational overhead. The code will be released in https://***/fukexue/McP-BERT. Copyright © 2022, The Authors. All rights reserved.

关键词： Large dataset

来源：评论

学校读者我要写书评

暂无评论

A single-loop accelerated extra-gradient difference algorithm with improved complexity bounds for constrained minimax optimization 23

A single-loop accelerated extra-gradient difference algorith...

引用

Proceedings of the 37th International Conference on Neural Information Processing Systems

作者： Yuanyuan Liu Fanhua Shang Weixin An Junhao Liu Hongying Liu Zhouchen Lin Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education School of Artificial Intelligence Xidian University China School of Computer Science and Technology College of Intelligence and Computing Tianjin University Medical College Tianjin University China and Peng Cheng Laboratory National Key Lab of General AI School of Intelligence Science and Technology Peking University and Institute for Artificial Intelligence Peking University and Peng Cheng Laboratory

In this paper, we propose a novel extra-gradient difference acceleration algorithm for solving constrained nonconvex-nonconcave (NC-NC) minimax problems. In particular, we design a new extra-gradient difference step to obtain an important quasi-cocoercivity property, which plays a key role to significantly improve the convergence rate in the constrained NC-NC setting without additional structural assumption. Then momentum acceleration is also introduced into our dual accelerating update step. Moreover, we prove that, to find an ε-stationary point of the function f, the proposed algorithm attains the complexity O(ε-2) in the constrained NC-NC setting, while the best-known complexity bound is Õ(ε-4), where Õ(·) hides logarithmic factors compared to Õ(·). As the special cases of the constrained NC-NC setting, the proposed algorithm can also obtain the same complexity Õ(ε-2) for both the nonconvex-concave (NC-C) and convex-nonconcave (C-NC) cases, while the best-known complexity bounds are Õ(ε-2.5) for the NC-C case and Õ(ε-4) for the C-NC case. For fair comparison with existing algorithms, we also analyze the complexity bound to find ε-stationary point of the primal function ϕ for the constrained NC-C problem, which shows that our algorithm can improve the complexity bound from Õ(ε-3) to Õ(ε-2). To the best of our knowledge, this is the first time that the proposed algorithm improves the best-known complexity bounds from Õ(ε-4) and Õ(ε-3) to Õ(ε-2) in both the NC-NC and NC-C settings.

关键词：

来源：评论

学校读者我要写书评

暂无评论

SETTP: Style Extraction and Tunable Inference via Dual-level Transferable Prompt Learning

arXiv

引用

arXiv 2024年

作者： Jin, Chunzhen Huang, Yongfeng Wang, Yaqi Cao, Peng Zaiane, Osmar Computer Science and Engineering Northeastern University Shenyang China Key Laboratory of Intelligent Computing in Medical Image of Ministry of Education Northeastern University Shenyang China Department of Computer Science and Engineering The Chinese University of Hong Kong Hong Kong National Frontiers Science Center for Industrial Intelligence and Systems Optimization Shenyang China Amii University of Alberta EdmontonAB Canada

Text style transfer, an important research direction in natural language processing, aims to adapt the text to various preferences but often faces challenges with limited resources. In this work, we introduce a novel method termed Style Extraction and Tunable Inference via Dual-level Transferable Prompt Learning (SETTP) for effective style transfer in low-resource scenarios. First, SETTP learns source style-level prompts containing fundamental style characteristics from high-resource style transfer. During training, the source style-level prompts are transferred through an attention module to derive a target style-level prompt for beneficial knowledge provision in low-resource style transfer. Additionally, we propose instance-level prompts obtained by clustering the target resources based on the semantic content to reduce semantic bias. We also propose an automated evaluation approach of style similarity based on alignment with human evaluations using ChatGPT-4. Our experiments across three resourceful styles show that SETTP requires only 1/20th of the data volume to achieve performance comparable to state-of-the-art (SOTA) methods. In tasks involving scarce data like writing style and role style, SETTP outperforms previous methods by 16.24%. Copyright © 2024, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Semivortex solitons and their excited states in spin-orbit-coupled binary bosonic condensates

引用

Physical Review E 2024年第6期109卷 064201-064201页

作者： Haiming Deng Jinqing Li Zhaopin Chen Yaohui Liu Dong Liu Chunzhi Jiang Chao Kong Boris A. Malomed School of Physics and Electronic-Electrical Engineering Microelectronics and Optoelectronics Technology Key Laboratory of Hunan Higher Education Hunan Engineering Research Center of Advanced Embedded Computing and Intelligent Medical Systems Physics Department and Solid-State Institute Department of Physical Electronics School of Electrical Engineering Faculty of Engineering Instituto de Alta Investigación

It is known that two-dimensional two-component fundamental solitons of the semivortex (SV) type, with vorticities (s+,s−)=(0,1) in their components, are stable ground states (GSs) in the spin-orbit-coupled (SOC) binary Bose-Einstein condensate with the contact self-attraction acting in both components, in spite of the possibility of the critical collapse in the system. However, excited states (ESs) of the SV solitons, with the vorticity set (s+,s−)=(S+,S++1) and S+=1,2,3,..., are unstable in the same system. We construct ESs of SV solitons in the SOC system with opposite signs of the self-interaction in the two components. The main finding is stability of the ES-SV solitons, with the extra vorticity (at least) up to S+=6. The threshold value of the norm for the onset of the critical collapse, Nthr, in these excited states is higher than the commonly known critical value, Nc≈5.85, associated with the single-component Townes solitons, Nthr increasing with the growth of S+. A velocity interval for stable motion of the GS-SV solitons is found too. The results suggest a solution for the challenging problem of the creation of stable vortex solitons with high topological charges.

关键词： Solitons

来源：评论

学校读者我要写书评

暂无评论

MPTA-Net: Multi-Scale Perception Network with Triple-View Attention for Polyp Segmentation 4

MPTA-Net: Multi-Scale Perception Network with Triple-View At...

引用

4th International Conference on Neural Networks, Information and Communication Engineering, NNICE 2024

作者： Chen, Zhihong Yao, Lisha Liu, Yue Qiu, Bingjiang Han, Chu Tang, Haoyang Du, Yujia Liu, Zaiyi Fang, Gang Guangzhou University Institute of Computing Science and Technology Guangzhou China Guangdong Prov. Key Laboratory of Artificial Intelligence in Medical Image Analysis and Application Guangzhou China South China University of Technology School of Medicine Guangzhou China Department of Radiology Guangzhou China Jiangxi College of Applied Technology School of Information Engineering Ganzhou China Guangdong Provincial People's Hospital Guangdong Academy of Sciences Guangdong Cardiovascular Institute Guangzhou China Southern Medical University Medical Research Institute Guangzhou China Kunming University of Science and Technology Faculty of Information Engineering and Automation Kunming China

ISBN: (纸本)9798350394375

Accurate segmentation of polyps in endoscopic images is pivotal for the early detection and diagnosis of colorectal cancer (CRC). Nonetheless, the varied appearance of polyp foregrounds and the intricate background interference significantly diminish the efficacy of pixel-level predictions. To address this issue, we propose a multi-scale perception network with triple-view attention (MPTA-Net) for polyp segmentation. The model primarily revolves around two core principles. The first integrates multiscale spatial information into the network's hierarchical encoding process to better preserve detailed polyp semantic information. The second introduces a triple-view attention mechanism that analyzes feature maps from three distinct perspectives to focus on and localize semantically relevant areas, thereby enhancing segmentation accuracy. Experimental results reveal that the proposed model outperforms existing mainstream segmentation models, effectively addressing under-segmentation, over-segmentation, and edge blurriness issues in polyp segmentation, thereby achieving better identification of challenging areas. © 2024 IEEE.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

An Attention-Based Signed Distance Field Estimation Method for Hand-Object Reconstruction

An Attention-Based Signed Distance Field Estimation Method f...

引用

Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW), IEEE Conference on

作者： Xinkang Zhang Xinhan Di Xiaokun Dai Xinrong Chen Academy for engineering & technology Fudan University shanghai China Bloo company shanghai China Academy for engineering & technology Shanghai Key Laboratory of Medical Image Computing and Computer Assisted Intervention Fudan University shanghai China

Joint reconstruction of hands and objects from monocular RGB images is a challenging task. In this work, we present a novel hybrid model for joint reconstruction of hands and objects. The model proposed consists of three key modules. Among them, multi-scale attention feature extractor is designed to enhance cross-scale in-formation extraction. Attention-based graph encoder can encode the graph information of the hands. Interacting attention module is applied to fuse information between hands and object. Test re-sults on ObMan dataset [6] show that our method outperforms the state-of-the-art method around 13.97% and 15.75% in $\mathrm{H}_{\text{se}}$ and $\mathrm{H}_{\text{je}}$ respectively.

关键词： Solid modeling Three-dimensional displays Fuses Estimation Virtual reality User interfaces Reconstruction algorithms

来源：评论

学校读者我要写书评

暂无评论

Megavoltage Computed Tomography (MVCT) Imaging Quality Improvement via Convolutional Neural Network

Megavoltage Computed Tomography (MVCT) Imaging Quality Impro...

引用

International Conference on image, Vision and intelligent Systems, ICIVIS 2021

作者： Zhao, Zengjing Dong, Jiwen Niu, Sijie Zhang, Yan Zhu, Jian School of Information Science and Engineering University of Jinan Jinan250022 China Shandong Provincial Key Laboratory of Network-Based Intelligent Computing Jinan250022 China Department of Radiation Oncology Physics and Technology Shandong Cancer Hospital affiliated to Shandong First Medical University Jinan250022 China

ISBN: (纸本)9789811669620

Clinical application of Megavoltage Computed Tomography (MVCT) is limited to three-dimensional visualization and position of the patient. The noise and low soft tissue contrast of MVCT make it difficult to offer accurate target volume delineation for adaptive radiotherapy and accurate image registration for image-guided radiotherapy, resulting in the less clinical application prospect of MVCT. In this work, we developed a deep learning-based approach to improve the MVCT image quality, so as to learn a mapping between MVCT images and paired planning kilovoltage CT (KVCT) images. The root mean squared error (RMSE), structural similarity (SSIM), and peak signal-to-noise ratio (PSNR) were used to quantify. Specifically, the PSNR and SSIM of the processed images improved to 27.94 ± 1.64 dB and 0.93 ± 0.01 compared with 24.67 ± 1.16 dB and 0.88 ± 0.02 in the original MVCT images, and the RMSE decreased from 23.5 ± 3.14HU to 16.34 ± 3.39HU. The processed images were obtained using the proposed method with less noise and higher soft tissue contrast in comparison with machine learning-based method. The processed MVCT increases the feasibility of depicting the anatomical delineation of tumor and organs at risk, and therefore enables the promising application of image-guided and adaptive helical TomoTherapy. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Weakly supervised alignment and registration of MR-CT for cervical cancer radiotherapy

arXiv

引用

arXiv 2024年

作者： Zhang, Jiahao Gu, Yin Sun, Deyu Gao, Yuhua Cui, Ming Zhang, Teng Ma, He College of Medicine and Biological Information Engineering Northeasten University Liaoning Shenyang110169 China Department of Radiation Oncology Gastrointestinal and Urinary and Musculoskeletal Cancer Cancer Hospital of Dalian University of Technology Liaoning Shenyang110110 China Department of Gynecological Radiotherapy Cancer Hospital of China Medical University Liaoning Shenyang110110 China Key Laboratory of Intelligent Computing in Medical Image Ministry of Education Shenyang110819 China

Cervical cancer is one of the leading causes of death in women, and brachytherapy is currently the primary treatment method. However, it is important to precisely define the extent of paracervical tissue invasion to improve cancer diagnosis and treatment options. The fusion of the information characteristics of both computed tomography (CT) and magnetic resonance imaging (MRI) modalities may be useful in achieving a precise outline of the extent of paracervical tissue invasion. Registration is the initial step in information fusion. However, when aligning multimodal images with varying depths, manual alignment is prone to large errors and is time-consuming. Furthermore, the variations in the size of the Region of Interest (ROI) and the shape of multimodal images pose a significant challenge for achieving accurate registration. In this paper, we propose a preliminary spatial alignment algorithm and a weakly supervised multimodal registration network. The spatial position alignment algorithm efficiently utilizes the limited annotation information in the two modal images provided by the doctor to automatically align multimodal images with varying depths. By utilizing aligned multimodal images for weakly supervised registration and incorporating pyramidal features and cost volume to estimate the optical flow, the results indicate that the proposed method outperforms traditional volume rendering alignment methods and registration networks in various evaluation metrics. This demonstrates the effectiveness of our model in multimodal image registration. Copyright © 2024, The Authors. All rights reserved.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

Sea Surface Temperature Prediction Approach Based on 3D CNN and LSTM with Attention Mechanism

Sea Surface Temperature Prediction Approach Based on 3D CNN ...

引用

International Conference on Advanced Communication Technology (ICACT)

作者： Baiyou Qiao Zhongqiang Wu Zhong Tang Gang Wu School of Computer Science and Engineering Northeastern University China Key Laboratory of Intelligent Computing in Medical Image Ministry of Education Northeastern University China

ISBN: (数字)9791188428069

ISBN: (纸本)9781665447027

Sea Surface Temperature (SST) is an important physical quantity of the ocean system. The accurate prediction of SST is essential for studying physical ocean phenomena and forecasting the ocean environment information. In this paper, a SST prediction approach based on 3-Diminsional Convolutional Neural Network (3D CNN) and Long Short-Term Memory (LSTM) network with attention mechanism is proposed, which considers the spatial correlation and temporal dependency of SST data. Firstly, the machine learning algorithm XGBoost is used to extract the long period time features of each SST data. Then the 3D CNN is used to capture the spatial correlations among SST field data composed of multiple observation points in a selected sea area, followed by the LSTM model to extract the time dependency features of the SST field time series data, and the attention mechanism is added to weight the output of each step of LSTM model to adjust the prediction results and improve the prediction accuracy of the approach. A series of experimental results show that the proposed approach has lower complexity, higher training efficiency and prediction accuracy, which is significantly better than the existing prediction models.

关键词： Sea surface Three-dimensional displays Correlation Computational modeling Predictive models Feature extraction Ocean temperature

来源：评论

学校读者我要写书评

暂无评论

A Multi-Objective Complex Network Pruning Framework Based on Divide-and-Conquer and Global Performance Impairment Ranking

SSRN

引用

SSRN 2024年

作者： Shang, Ronghua Zhu, Songling Zhang, Weitong Xu, Songhua Li, Yangyang Key Laboratory of Intelligent Perception and Image Understanding Ministry of Education School of Artificial Intelligence Xidian University Shaanxi Province Xi’an710071 China Department of Health Management Institute of Medical Artificial Intelligence The Second Affiliated Hospital of Xi’an Jiaotong University Xi’an China

Model compression plays a vital role in the practical deployment of deep neural networks (DNNs), and evolutionary multi-objective (EMO) pruning is an essential tool in balancing the compression rate and performance of the DNNs. However, due to its populationbased nature, EMO pruning suffers from the complex optimization space and the resource-intensive structure verification process. To this end, a multi-objective complex network pruning framework based on divide-and-conquer and global performance impairment ranking (EMO-DIR) is proposed in this paper. Firstly, a divide-and-conquer EMO network pruning method is proposed, which decomposes the complex task of EMO pruning on the entire network into easier sub-tasks on multiple sub-networks. Because the network structure to be optimized and tested becomes smaller, this decomposition narrows the pruning optimization space and reduces resource consumption. However, the decomposition also breaks the cooperation between different sub-networks and hinders designing a joint pruning scheme. So, a sub-network training method based on cross-network constraints is designed, which could bridge independent EMO pruning sub-tasks, allowing them to collaborate better and improving the overall performance of the pruned network. Then, a multiple sub-networks joint pruning method based on EMO is proposed. This method combines the Pareto Fronts from EMO pruning results on multiple sub-networks through global performance impairment ranking to design a joint pruning scheme. The rich experiments on CIFAR-10/100 are conducted. EMO-DIR achieves a comparable performance with the state-of-the-art pruning methods. © 2024, The Authors. All rights reserved.

关键词： image classification

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：