检索结果-内蒙古大学图书馆

arXiv 2021年

作者： Fu, Kexue Liu, Shaolei Luo, Xiaoyuan Wang, Manning Digital Medical Research Center School of Basic Medical Science Fudan University Shanghai Key Lab of Medical Image Computing and Computer Assisted Intervention China

3D point cloud registration is a fundamental problem in computer vision and robotics. There has been extensive research in this area, but existing methods meet great challenges in situations with a large proportion of outliers and time constraints, but without good transformation initialization. Recently, a series of learning-based algorithms have been introduced and show advantages in speed. Many of them are based on correspondences between the two point clouds, so they do not rely on transformation initialization. However, these learning-based methods are sensitive to outliers, which lead to more incorrect correspondences. In this paper, we propose a novel deep graph matching-based framework for point cloud registration. Specifically, we first transform point clouds into graphs and extract deep features for each point. Then, we develop a module based on deep graph matching to calculate a soft correspondence matrix. By using graph matching, not only the local geometry of each point but also its structure and topology in a larger range are considered in establishing correspondences, so that more correct correspondences are found. We train the network with a loss directly defined on the correspondences, and in the test stage the soft correspondences are transformed into hard one-to-one correspondences so that registration can be performed by singular value decomposition. Furthermore, we introduce a transformer-based method to generate edges for graph construction, which further improves the quality of the correspondences. Extensive experiments on registering clean, noisy, partial-to-partial and unseen category point clouds show that the proposed method achieves state-of-the-art performance. The code will be made publicly available at https://***/fukexue/RGM. Copyright © 2021, The Authors. All rights reserved.

关键词： Surface measurement

来源：评论

学校读者我要写书评

暂无评论

Boosting 3D Point Cloud Registration by Transferring Multi-modality Knowledge

Boosting 3D Point Cloud Registration by Transferring Multi-m...

引用

IEEE International Conference on Robotics and Automation (ICRA)

作者： Mingzhi Yuan Xiaoshui Huang Kexue Fu Zhihao Li Manning Wang 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 Shanghai artificial intelligence Lab. China

The recent multi-modality models have achieved great performance in many vision tasks because the extracted features contain the multi-modality knowledge. However, most of the current registration descriptors have only concentrated on local geometric structures. This paper proposes a method to boost point cloud registration accuracy by transferring the multi-modality knowledge of pre-trained multi-modality model to a new descriptor neural network. Different to the previous multi-modality methods that requires both modalities, the proposed method only requires point clouds during inference. Specifically, we propose an ensemble descriptor neural network combining pre-trained sparse convolution branch and a new point-based convolution branch. By fine-tuning on a single modality data, the proposed method achieves new state-of-the-art results on 3DMatch and competitive accuracy on 3DLoMatch and KITTI. The code and the trained model will be released at https://***/phdymz/***.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Improvement of ultrasound contrast imaging with adaptive beamformer based on pulse inversion plane wave transmission

Improvement of ultrasound contrast imaging with adaptive bea...

引用

URSI Asia-Pacific Radio Science Conference (URSI AP-RASC)

作者： Yurong Huang Jinhua Yu Yuanyuan Wang Shuying Li Yi Chen Department of Electronic Engineering Fudan University Shanghai China The key laboratory of medical imaging computing and computer assisted intervention of shanghai Shanghai China

Study of ultrasound contrast imaging is of great importance in recent years thanks to its wide application in ultrasonic diagnosis. Plane wave imaging has been considered as a good method with high frame rate and low mechanical index. High frame rate can improve the temporal resolution of contrast imaging. Meanwhile, low mechanical index is essential to contrast imaging since high mechanical index in traditional B-mode imaging could break micro-bubbles easily. However, plane wave contrast imaging suffers from low signal-to-noise ratio (SNR) and contrast (CR) which leads to poor image quality. In this paper, adaptive beamformer and pulse inversion (PI) technique were combined to solve the shortcomings of the plane wave contrast imaging. For the adaptive beamformer, eigenspace based minimum variance (ESBMV) with coherence factor (ESBMVCF) was proposed. Experimental results show that ESBMVCF offers an enhancement of 1.1dB, 13dB and 0.8dB in CR, contrast-to-tissue ratio (CTR) and contrast-to-noise ratio (CNR) respectively compared with minimum variance and coherence factor (MVCF). Besides, PI offers 1.5dB, 15dB and 15dB enhancement in CR, CTR and CNR compared with traditional pulse-echo imaging.

关键词： Radar imaging Meteorology Imaging Meteorological radar image resolution Radar antennas

来源：评论

学校读者我要写书评

暂无评论

Adaptive scaled Wiener postfilter beamformer for ultrasound imaging

Adaptive scaled Wiener postfilter beamformer for ultrasound ...

引用

URSI Asia-Pacific Radio Science Conference (URSI AP-RASC)

作者： Saeid Aliabadi Yuanyuan Wang Jinhua Yu Department of Electronic Engineering Fudan University Shanghai China The key laboratory of medical imaging computing and computer assisted intervention of shanghai Shanghai China

In ultrasound imaging Wiener postfilter and coherent factor (CF) concepts were combined with minimum variance (MV) and eigenspace-based MV (ESBMV) beamformers to future suppress sidelobes and to improve the imaging contrast and the spatial resolution. CF, which can be considered as a scaled Wiener postfilter with high noise power weighting, provides high imaging resolution. The high resolution, however, is achieved with the contrast as a sacrifice. On other hand, Wiener postfilter, provides better signal intensity restoration and imaging contrast than CF but with lower imaging resolution. To pursue the optimal imaging contrast of Wiener postfilter and the high resolution of CF at the same time, an adaptive scaled Wiener postfilter beamformer (ASC-Wiener) is proposed. In proposed method, the noise power weighting of Wiener postfilter is adjusted adaptively so that related to the output SNR inversely needless to estimate the input SNR. Simulation and real data experiment for target point imaging show the high performance of the proposed beamformer in terms of the imaging resolution and sidelobes suppression. Cyst phantom imaging also shows that the ASC-Wiener preserves the speckle intensity and improves the imaging contrast. The contrast ratio (CR) and the contrast-to-noise ratio (CNR) are improved by 6.2/0.35 dB over ESBMV-Wiener and by 20.7/0.65 over ESBMV-CF. These results solidly prove that the imaging contrast and the imaging resolution are improved significantly and simultaneously.

关键词： Imaging Radar imaging image resolution Signal resolution Meteorology Ultrasonic imaging

来源：评论

学校读者我要写书评

暂无评论

Chest-Diffusion: A Light-Weight Text-to-image Model for Report-to-CXR Generation

Chest-Diffusion: A Light-Weight Text-to-Image Model for Repo...

引用

IEEE International Symposium on Biomedical Imaging

作者： Peng Huang Xue Gao Lihong Huang Jing Jiao Xiaokang Li Yuanyuan Wang Yi Guo Department of Electronic Engineering Fudan University Shanghai China Key Laboratory of Medical Imaging Computing and Computer Assisted Intervention of Shanghai Shanghai China

ISBN: (数字)9798350313338

ISBN: (纸本)9798350313345

Text-to-image generation has important implications for generation of diverse and controllable images. Several attempts have been made to adapt Stable Diffusion (SD) to the medical domain. However, the large distribution difference between medical reports and natural texts, as well as high computational complexity in common stable diffusion limit the authenticity and feasibility of the generated medical images. To solve above problems, we propose a novel light-weight transformer-based diffusion model learning framework, Chest-Diffusion, for report-to-CXR generation. Chest-Diffusion employs a domain-specific text encoder to obtain accurate and expressive text features to guide image generation, improving the authenticity of the generated images. Meanwhile, we introduce a light-weight transformer architecture as the denoising model, reducing the computational complexity of the diffusion model. Experiments demonstrate that our Chest-Diffusion achieves the lowest FID score 24.456, under the computation budget of 118.918 GFLOPs, which is nearly one-third of the computational complexity of SD.

关键词： Accuracy Computational modeling Noise reduction Text to image computer architecture Transformers Diffusion models

来源：评论

学校读者我要写书评

暂无评论

The theoretical and experimental analysis of ultrasound contrast agent micro bubbles: The Sonovue

The theoretical and experimental analysis of ultrasound cont...

引用

International Conference on Biomedical Engineering and Informatics (BMEI)

作者： Shuying Li Yurong Huang Yi Cheng Jinhua Yu Yuanyuan Wang Department of Electronic Engineering Fudan University Shanghai China The key laboratory of medical imaging computing and computer assisted intervention of shanghai China

Sonovue is a kind of ultrasound contrast agent (UCA) with a stable shell made by lipid, within which is SF6. Due to the specific acoustic characteristics, UCA has been widely used in ultrasound diagnosis and therapy. In some cases, the 2nd harmonic wave of the scattered signal should be strong enough to help form a high quality image. To better utilize the bubbles to get enough intensity of the 2nd harmonic wave of the scattered signal, numerical simulation and corresponding experiments based on the several models proposed before have been analyzed to study the behavior of the bubbles. In this paper, we use the ratio of the second harmonic wave and fundamental wave to evaluate the oscillation of the bubbles. Besides, we have compared different simulation models and designed corresponding platform of phantom study. Finally, we find that the second harmonic wave will get a remarkable growth when the driving pressure is under 150KPa in the simulation or under 200KPa in the experiment.

关键词： Mathematical model Harmonic analysis Numerical models Acoustics Liquids Phantoms Ultrasonic imaging

来源：评论

学校读者我要写书评

暂无评论

Fusionmlp: A Mlp-Based Unified image Fusion Framework

SSRN

引用

SSRN 2024年

作者： Liu, Shaolei Li, Shiman Qu, Linhao Wang, Manning Song, Zhijian Digital Medical Research Center School of Basic Medical Science Fudan University Shanghai200032 China Shanghai Key Lab of Medical Image Computing and Computer Assisted Intervention China

Due to the powerful feature representation capacity, deep learning-based image fusion methods have improved the fusion results for better information integration. However, some inherent limitations in convolutional neural networks (CNN) and Transformer networks impede further performance improvement. Specifically, CNN-based methods have difficulties in mining global context information;Transformer-based methods are unable to adapt to different image sizes and the computational complexity grows exponentially with the image size. To overcome these problems, we propose the first MLP-based multi-scene image fusion framework, denoted as FusionMLP. In addition, we propose a novel loss function combining the spatial and frequency domain information for model training. Extensive experiments show that FusionMLP not only outperform SOTA multi-scene image fusion methods, but also outperform SOTA fusion algorithms for specific scenes. © 2024, The Authors. All rights reserved.

关键词： image fusion

来源：评论

学校读者我要写书评

暂无评论

Boosting 3D Point Cloud Registration by Transferring Multi-modality Knowledge

arXiv

引用

arXiv 2023年

作者： Yuan, Mingzhi Huang, Xiaoshui Fu, Kexue Li, Zhihao Wang, Manning The Digital Medical Research Center School of Basic Medical Science Fudan University Shanghai200032 China The Shanghai Key Laboratory of Medical Image Computing and Computer Assisted Intervention Shanghai200032 China The Shanghai artificial intelligence Lab China

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Fusionmlp: A Mlp-Based Unified image Fusion Framework

SSRN

引用

SSRN 2023年

作者： Liu, Shaolei Qu, Linhao Wang, Manning Song, Zhijian Digital Medical Research Center School of Basic Medical Science Fudan University Shanghai200032 China Shanghai Key Lab of Medical Image Computing and Computer Assisted Intervention China

Due to the powerful feature representation capacity, deep learning-based image fusion methods have improved the fusion results for better information integration. However, some inherent limitations in convolutional neural networks (CNN) and Transformer networks impede further performance improvement. Specifically, CNN-based methods have difficulties in mining global context information;Transformer-based methods are unable to adapt to different image sizes and the computational complexity grows exponentially with the image size. To overcome these problems, we propose the first MLPbased multi-scene image fusion framework, denoted as FusionMLP. In addition, we propose a novel loss function combining the spatial and frequency domain information for model training. Extensive experiments show that FusionMLP not only outperform SOTA multi-scene image fusion methods, but also outperform SOTA fusion algorithms for specific scenes. Codes will be released at https://***/slliuEric/FusionMLP. © 2023, The Authors. All rights reserved.

关键词： image fusion

来源：评论

学校读者我要写书评

暂无评论

Reducing Domain Gap in Frequency and Spatial domain for Cross-modality Domain Adaptation on medical image Segmentation

arXiv

引用

arXiv 2022年

作者： Liu, Shaolei Yin, Siqi Qu, Linhao Wang, Manning Digital Medical Research Center School of Basic Medical Science Fudan University Shanghai200032 China Shanghai Key Lab of Medical Image Computing and Computer Assisted Intervention China

Unsupervised domain adaptation (UDA) aims to learn a model trained on source domain and performs well on unlabeled target domain. In medical image segmentation field, most existing UDA methods depend on adversarial learning to address the domain gap between different image modalities, which is ineffective due to its complicated training process. In this paper, we propose a simple yet effective UDA method based on frequency and spatial domain transfer under multi-teacher distillation framework. In the frequency domain, we first introduce non-subsampled contourlet transform for identifying domain-invariant and domain-variant frequency components (DIFs and DVFs), and then keep the DIFs unchanged while replacing the DVFs of the source domain images with that of the target domain images to narrow the domain gap. In the spatial domain, we propose a batch momentum update-based histogram matching strategy to reduce the domain-variant image style bias. Experiments on two cross-modality medical image segmentation datasets (cardiac, abdominal) show that our proposed method achieves superior performance compared to state-of-the-art methods. Copyright © 2022, The Authors. All rights reserved.

关键词： Distillation

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：