检索结果-内蒙古大学图书馆

Attentional Ptycho-Tomography(APT)for three-dimensional nanoscale X-ray imaging with minimal data acquisition and computation time

引用

Light(science & Applications) 2023年第6期12卷 1127-1140页

作者： Iksung Kang Ziling Wu Yi Jiang Yudong Yao Junjing Deng Jeffrey Klug Stefan Vogt George Barbastathis Department of Electrical Engineering and Computer Science Massachusetts Institute of TechnologyCambridgeMA 02139USA Department of Mechanical Engineering Massachusetts Institute of TechnologyCambridgeMA 02139USA Argonne National Laboratory LemontIL 60439USA Singapore-MIT Alliance for Research and Technology(SMART)Centre 1 CREATE WaySingapore 138602Singapore Present address:Department of Molecular and Cell Biology University of CaliforniaBerkeleyCA 94720USA Present address:Singapore-MIT Alliance for Research and Technology(SMART)Centre 1 CREATE WaySingapore 138602Singapore Present address:Center for Transformative Science ShanghaiTech University201210 ShanghaiChina

Noninvasive X-ray imaging of nanoscale three-dimensional objects,such as integrated circuits(ICs),generally requires two types of scanning:ptychographic,which is translational and returns estimates of the complex electromagnetic field through the IC;combined with a tomographic scan,which collects these complex field projections from multiple ***,we present Attentional Ptycho-Tomography(APT),an approach to drastically reduce the amount of angular scanning,and thus the total acquisition *** is machine learning-based,utilizing axial self-Attention for Ptycho-Tomographic *** is trained to obtain accurate reconstructions of the ICs,despite the incompleteness of the *** training process includes regularizing priors in the form of typical patterns found in IC interiors,and the physics of X-ray propagation through the *** show that APT with×12 reduced angles achieves fidelity comparable to the gold standard Simultaneous Algebraic Reconstruction Technique(SART)with the original set of *** using the same set of reduced angles,then APT also outperforms Filtered Back Projection(FBP),Simultaneous Iterative Reconstruction Technique(SIRT)and *** time needed to compute the reconstruction is also reduced,because the trained neural network is a forward operation,unlike the iterative nature of these *** experiments show that,without loss in quality,for a 4.48×93.2×3.92µm^(3) IC(≃6×10^(8) voxels),APT reduces the total data acquisition and computation time from 67.96 h to 38 *** expect our physics-assisted and attention-utilizing machine learning framework to be applicable to other branches of nanoscale imaging,including materials science and biological imaging.

关键词： utilizing computation interior

来源：评论

学校读者我要写书评

暂无评论

Equivariant frames and the impossibility of continuous canonicalization 24

Equivariant frames and the impossibility of continuous canon...

引用

Proceedings of the 41st International Conference on Machine Learning

作者： Nadav Dym Hannah Lawrence Jonathan W Siegel Faculty of Mathematics Faculty of Computer Science Technion Israel Department of Electrical Engineering and Computer Science MIT MA Department of Mathematics Texas A&M University TX

Canonicalization provides an architecture-agnostic method for enforcing equivariance, with generalizations such as frame-averaging recently gaining prominence as a lightweight and flexible alternative to equivariant architectures. Recent works have found an empirical benefit to using probabilistic frames instead, which learn weighted distributions over group elements. In this work, we provide strong theoretical justification for this phenomenon: for commonly-used groups, there is no efficiently computable choice of frame that preserves continuity of the function being averaged. In other words, unweighted frame-averaging can turn a smooth, non-symmetric function into a discontinuous, symmetric function. To address this fundamental robustness problem, we formally define and construct weighted frames, which provably preserve continuity, and demonstrate their utility by constructing efficient and continuous weighted frames for the actions of SO(d), O(d), and Sn on point clouds.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Three-dimensional nanoscale reduced-angle ptycho-tomographic imaging with deep learning(RAPID)

引用

eLight 2023年第1期3卷 198-210页

作者： Ziling Wu Iksung Kang Yudong Yao Yi Jiang Junjing Deng Jeffrey Klug Stefan Vogt George Barbastathis Department of Mechanical Engineering Massachusetts Institute of TechnologyCambridgeMA 02139USA Department of Electrical Engineering and Computer Science Massachusetts Institute of TechnologyCambridgeMA 02139USA Argonne National Laboratory LemontIL 60439USA Singapore-MIT Alliance for Research and Technology(SMART)Centre Singapore 138602Singapore Present Address:Singapore-MIT Alliance for Research and Technology(SMART)Centre Singapore 138602Singapore Present Address:University of California BerkeleyCA 94720USA

X-ray ptychographic tomography is a nondestructive method for three dimensional(3D)imaging with nanometer-sized resolvable *** size of the volume that can be imaged is almost arbitrary,limited only by the penetration depth and the available scanning *** we present a method that rapidly accelerates the imaging operation over a given volume through acquiring a limited set of data via large angular reduction and compensating for the resulting ill-posedness through deeply learned *** proposed 3D reconstruction method“RAPID”relies initially on a subset of the object measured with the nominal number of required illumination angles and treats the reconstructions from the conventional two-step approach as ground *** is then trained to reproduce equal fidelity from much fewer *** training,it performs with similar fidelity on the hitherto unexamined portions of the object,previously not shown during training,with a limited set of *** our experimental demonstration,the nominal number of angles was 349 and the reduced number of angles was 21,resulting in a×140 aggregate speedup over a volume of 4.48×93.18×3.92μm^(3) and with(14 nm)^(3) feature size,i.e.-10^(8) ***’s key distinguishing feature over earlier attempts is the incorporation of atrous spatial pyramid pooling modules into the deep neural network framework in an anisotropic *** found that adjusting the atrous rate improves reconstruction fidelity because it expands the convolutional kernels’range to match the physics of multi-slice ptychography without significantly increasing the number of parameters.

关键词： X-ray ptychographic tomography Deep learning Reduced-angle Rapid imaging Atrous spatial pyramid pooling Anisotropic

来源：评论

学校读者我要写书评

暂无评论

An Efficient Approach of Brain Tumor Detection & Extraction using BWT with Auto Enhance Technique

引用

Procedia computer science 2025年 258卷 4105-4116页

作者： Nilesh Bhaskarrao Bahadure Sudhanshu Gonge Jagdish Chandra Patni Bhoomi Shah Nagrajan Raju Prasenjeet D. Patil Department of Computer Science and Engineering School of Technology GSFC University Vadodara Gujarat India Department of Computer Science and Engineering Symbiosis Institute of Technology Symbiosis International (Deemed University) Pune Maharashtra India Department of Computer Science and Engineering Alliance School of Advanced Computing Alliance University Bengaluru India Department of Electrical and Electronics Engineering SASTRA University Thanjavur Tamil Nadu India Department of Computer Science and Engineering MIT ADT University Pune India

Brain tumor detection and extraction of infected areas with utmost accuracy is essential in medical image analysis. Despite ex- cessive research, segmentation operation for brain tumor detection remains an exigent issue because medical images have diverse image contents, non-uniformity object texture, occlusion, image noise, and other factors that consistently influence and affect per- formance analysis. This paper introduces the most efficient brain tumor detection and extraction technique, using Berkely Wavelet Transformation-based segmentation with auto-enhancement techniques. If the input images are too dark or their intensity is not good enough, the auto-enhance technique will automatically adjust the image quality. The extensive experimental analysis and sim- ulation, with a comparison of the results with other state-of-the-art techniques available, prove the significance and effectiveness of the proposed algorithm. The primary intent of this paper is to develop a simple but performance-oriented algorithm for brain tumor detection and extraction, which is suitable for most imaging modalities and gives the best performance even if images are of poor quality.

关键词： Berkeley wavelet transformation enhancement magnetic resonance imaging (MRI) segmentation morphology

来源：评论

学校读者我要写书评

暂无评论

On the hardness of learning under symmetries

arXiv

引用

arXiv 2024年

作者： Kiani, Bobak T. Le, Thien Lawrence, Hannah Jegelka, Stefanie Weber, Melanie John A. Paulson School of Engineering and Applied Sciences Harvard Department of Electrical Engineering and Computer Science MIT United States Department of Electrical Engineering and Computer Science MIT United States School of Computation Information and Technology TU Munich Germany John A. Paulson School of Engineering and Applied Sciences Harvard United States

We study the problem of learning equivariant neural networks via gradient descent. The incorporation of known symmetries ("equivariance") into neural nets has empirically improved the performance of learning pipelines, in domains ranging from biology to computer vision. However, a rich yet separate line of learning theoretic research has demonstrated that actually learning shallow, fully-connected (i.e. non-symmetric) networks has exponential complexity in the correlational statistical query (CSQ) model, a framework encompassing gradient descent. In this work, we ask: are known problem symmetries sufficient to alleviate the fundamental hardness of learning neural nets with gradient descent? We answer this question in the negative. In particular, we give lower bounds for shallow graph neural networks, convolutional networks, invariant polynomials, and frame-averaged networks for permutation subgroups, which all scale either superpolynomially or exponentially in the relevant input dimension. Therefore, in spite of the significant inductive bias imparted via symmetry, actually learning the complete classes of functions represented by equivariant neural networks via gradient descent remains hard. © 2024, CC BY.

关键词： Gradient methods

来源：评论

学校读者我要写书评

暂无评论

On the Use of Deep Learning for Three-Dimensional Computational Imaging 37

On the Use of Deep Learning for Three-Dimensional Computatio...

引用

Practical Holography XXXVII: Displays, Materials, and Applications 2023

作者： Barbastathis, George Pang, Subeen Kang, Iksung Wu, Ziling Liu, Zhiguang Guo, Zhen Zhang, Fucai Department of Mechanical Engineering Massachusetts Institute of Technology United States Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology United States Singapore-MIT Alliance for Research & Technology Centre Massachusetts Institute of Technology United States Department of Molecular and Cell Biology University of California Berkeley United States Department of Electrical Engineering Southern University of Science and Technology China

ISBN: (纸本)9781510659957

Deep learning has proven to be an efficient and robust method for many computational imaging systems.2 The advantages of machine learning, as a rule, are that it is fast-at least in its supervised form after training is complete-and seems exceedingly effective in capturing regularizing priors. Here, we focus the discussion on non-invasive three-dimensional (3D) object reconstruction. One then faces the additional dilemma of choosing the appropriate model of light-matter interaction inside the specimen, i.e. the forward operator. We describe the three stages of approximation that are applicable: weak scattering with weak diffraction (also known as the Radon transform), weak scattering with strong diffraction, and strong scattering. We then overview machine learning approaches for the various models, and glance at the consequences of oversimplifying the forward operator choice. © 2023 SPIE.

关键词： Diffraction

来源：评论

学校读者我要写书评

暂无评论

Analytic Design of Flat-Wire Inductors for High-Current and Compact DC-DC Converters

Analytic Design of Flat-Wire Inductors for High-Current and ...

引用

Annual IEEE Conference on Applied Power Electronics Conference and Exposition (APEC)

作者： Sajjad Mohammadi James L. Kirtley Alireza Namadmalan Electrical Engineering and Computer Science MIT Cambridge USA Power Electronics & Magnetics Lab Bourns Electronics Ireland (BEI) Cork Ireland

ISBN: (数字)9798331516116

ISBN: (纸本)9798331516123

This paper presents analytic study and design considerations of flat wire inductors with distributed gaps for high-power and compact DC-DC Converters. The focus is eddy-current loss components within the conductors due to fringing and leakage fluxes. A magnetic equivalent circuit (MEC) is proposed in which eddy currents are modeled by MMFs opposing the primary flux as well as frequency dependent reluctances, which finally leads to a frequency dependent inductance describing the behavior of the inductor at high frequencies. Three formulations for DC resistance depending on the required accuracy are developed. Calculations of the AC resistance based on vector potential obtained from FEM are provided. To provide an insight into the optimized design of such inductors, components of the magnetic flux and induced eddy currents along with sensitivity of the main inductor quantities such as DCR, ESR, loss components and inductance values to the design parameters are investigated. Finally, an inductor is prototyped and experimentally tested to verify the design.

关键词： Resistance Magnetic flux Inductance Sensitivity Finite element analysis Wire Inductors Eddy currents Magnetic circuits Equivalent circuits

来源：评论

学校读者我要写书评

暂无评论

Position: scarce resource allocations that rely on machine learning should be randomized 24

Position: scarce resource allocations that rely on machine l...

引用

Proceedings of the 41st International Conference on Machine Learning

作者： Shomik Jain Kathleen Creel Ashia Wilson Institute for Data Systems and Society MIT Department of Philosophy & Religion and Khoury College of Computer Sciences Northeastern University Institute for Data Systems and Society and Department of Electrical Engineering and Computer Science MIT

Contrary to traditional deterministic notions of algorithmic fairness, this paper argues that fairly allocating scarce resources using machine learning often requires randomness. We address why, when, and how to randomize by proposing stochastic procedures that more adequately account for all of the claims that individuals have to allocations of social goods or opportunities.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Allocation Multiplicity: Evaluating the Promises of the Rashomon Set

arXiv

引用

arXiv 2025年

作者： Jain, Shomik Wang, Margaret Creel, Kathleen Wilson, Ashia Institute for Data Systems and Society MIT United States Department of Electrical Engineering and Computer Science MIT United States Department of Philosophy and Religion Khoury College of Computer Sciences Northeastern University United States

The Rashomon set of equally-good models promises less discriminatory algorithms, reduced outcome homogenization, and fairer decisions through model ensembles or reconciliation. However, we argue from the perspective of allocation multiplicity that these promises may remain unfulfilled. When there are more qualified candidates than resources available, many different allocations of scarce resources can achieve the same utility. This space of equal-utility allocations may not be faithfully reflected by the Rashomon set, as we show in a case study of healthcare allocations. We attribute these unfulfilled promises to several factors: limitations in empirical methods for sampling from the Rashomon set, the standard practice of deterministically selecting individuals with the lowest risk, and structural biases that cause all equally-good models to view some qualified individuals as inherently risky. © 2025, CC BY.

关键词： Consensus algorithm

来源：评论

学校读者我要写书评

暂无评论

RNAFlow: RNA structure & sequence design via inverse folding-based flow matching 24

RNAFlow: RNA structure & sequence design via inverse folding...

引用

Proceedings of the 41st International Conference on Machine Learning

作者： Divya Nori Wengong Jin Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology Cambridge MA Broad Institute of MIT and Harvard Cambridge MA

The growing significance of RNA engineering in diverse biological applications has spurred interest in developing AI methods for structure-based RNA design. While diffusion models have excelled in protein design, adapting them for RNA presents new challenges due to RNA's conformational flexibility and the computational cost of fine-tuning large structure prediction models. To this end, we propose RNAFlow, a flow matching model for protein-conditioned RNA sequencestructure design. Its denoising network integrates an RNA inverse folding model and a pretrained RosettaFold2NA network for generation of RNA sequences and structures. The integration of inverse folding in the structure denoising process allows us to simplify training by fixing the structure prediction network. We further enhance the inverse folding model by conditioning it on inferred conformational ensembles to model dynamic RNA conformations. Evaluation on protein-conditioned RNA structure and sequence generation tasks demonstrates RNAFlow's advantage over existing RNA design methods.

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：