检索结果-内蒙古大学图书馆

IK-Geo: Unified Robot Inverse Kinematics Using Subproblem Decomposition

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Elias, Alexander J. Wen, John T. Department of Electrical Computer and Systems Engineering Rensselaer Polytechnic Institute TroyNY12180 United States

This paper presents the open-source robot inverse kinematics (IK) solver IK-Geo, the fastest general IK solver based on published literature. In this unifying approach, IK for any 6-DOF all-revolute (6R) manipulator is decomposed into six canonical geometric subproblems solved by intersecting circles with other geometric objects. We present new efficient and singularity-robust solutions to these subproblems using geometric and linear algebra methods. IK-Geo finds all IK solutions including singular solutions and sometimes least-squares solutions by solving for subproblem solutions in all cases, including in a continuous and sometimes least-squares sense when a solution does not exist. Robots are classified into kinematic families based on cases of intersecting or parallel joint axes, and robots in the same family use the same IK algorithm. 6R robots with three intersecting or parallel axes are solved in closed form, and all solutions are found exactly without iteration. Other 6R robots are efficiently solved by searching for zeros of an error function of one or two joint angles. The subproblem and IK solutions are easy to understand, implement, test, and modify, meaning this method is readily ported to new languages and environments. We connect our geometric method with less efficient but more robust polynomial-based methods: rather than using search, subproblems and error functions may be written in terms of the tangent half-angle of one joint. This results in a system of multivariate polynomial equations from which the univariate polynomial with zeros corresponding to IK solutions is readily derived. © 2022, CC BY.

关键词： Inverse kinematics

THz detection and amplification using plasmonic Field Effect Transistors driven by DC drain currents

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Zhang, Yuhui Shur, Michael Department of Electrical Computer and Systems Engineering Rensselaer Polytechnic Institute TroyNY12180 United States

We report on the numerical and theoretical results of sub-THz and THz detection by a current-driven InGaAs/GaAs plasmonic Field-Effect Transistor (TeraFET). New equations are developed to account for the channel length dependence of the drain voltage and saturation current. Numerical simulation results demonstrate that the effect of drain bias current on the source-to-drain response voltage (dU) varies with the device channel length. In a long-channel TeraFET where plasmonic oscillations cannot reach the drain, dU is always positive and rises rapidly with increasing drain current. For a short device in which plasmonic oscillations reach the drain, the current-induced nonuniform electric field leads to a negative response, agreeing with previous observations. At negative dU, the amplitude of the small-signal voltage at the drain side becomes larger than that at the source side. Thus, the device effectively serves as a THz amplifier in this condition. Under the resonant mode, the negative response can be further amplified near the resonant peaks. A new expression of dU is proposed to account for this resonant effect. Based on those expressions, a current-driven TeraFET spectrometer is proposed. The ease of implementation and simplified calibration procedures make it competitive or superior compared with other TeraFET-based spectrometers. © 2022, CC BY-NC-ND.

关键词： Drain current

Joint Beam Search Integrating CTC, Attention, and Transducer Decoders

学校读者我要写书评

暂无评论

IEEE Transactions on Audio, Speech and Language Processing

IEEE Transactions on Audio, Speech and Language Processing 2025年 33卷 598-612页

作者： Yui Sudo Muhammad Shakeel Yosuke Fukumoto Brian Yan Jiatong Shi Yifan Peng Shinji Watanabe Research Division Honda Research Institute Japan Company Ltd. Wakō-shi Saitama Japan Language Technologies Institute Carnegie Mellon University Pittsburgh PA USA Department of Electrical and Computer Engineering Carnegie Mellon University Pittsburgh PA USA

End-to-end automatic speech recognition (E2E-ASR) can be classified by its decoder architectures, such as connectionist temporal classification (CTC), recurrent neural network transducer (RNN-T), attention-based encoder-decoder, and Mask-CTC models. Each decoder architecture has advantages and disadvantages, leading practitioners to switch between these different models depending on application requirements. Instead of building separate models, we propose a joint modeling scheme where four decoders (CTC, RNN-T, attention, and Mask-CTC) share the same encoder – we refer to this as 4D modeling. The 4D model is trained jointly, which will bring model regularization and maximize the model robustness thanks to their complementary properties. To efficiently train the 4D model, we introduce a two-stage training strategy that stabilizes the joint training. In addition, we propose three novel joint beam search algorithms by combining three decoders (CTC, RNN-T, and attention) to further improve performance. These three beam search algorithms differ in which decoder is used as the primary decoder. We carefully evaluate the performance and computational tradeoffs associated with each algorithm. Experimental results demonstrate that the jointly trained 4D model outperforms the E2E-ASR models trained with only one individual decoder. Furthermore, we demonstrate that the proposed joint beam search algorithm outperforms the previously proposed CTC/attention decoding.

关键词： Decoding Hidden Markov models Computational modeling Training Computational efficiency Transducers Inference algorithms Accuracy Vectors Translation

FEDERATED LEARNING IN HEALTHCARE: MODEL MISCONDUCTS, SECURITY, CHALLENGES, APPLICATIONS, AND FUTURE RESEARCH DIRECTIONS-A SYSTEMATIC REVIEW

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Ali, Md Shahin Tasnim, Lamia Ahsan, Md Manjurul Afrin, Sadia Biswas, Koushik Hossain, Md Maruf Ahmed, Md Mahfuz Islam, Md Khairul Hashan, Ronok Raman, Shivakumar Department of Biomedical Engineering Islamic University Kushtia7003 Bangladesh Department of Food and Nutrition Government College of Applied Human Science Azimpur Dhaka1205 Bangladesh Department of Industrial and Systems Engineering University of Oklahoma Norman OK73071 United States Department of Computer Science and Engineering IIIT Delhi New Delhi Delhi110020 India Department of Electrical and Electronic Engineering Islamic University Kushtia7003 Bangladesh

Data privacy has become a major concern in healthcare due to the increasing digitization of medical records and data-driven medical research. Protecting sensitive patient information from breaches and unauthorized access is critical, as such incidents can have severe legal and ethical complications. Federated Learning (FL) addresses this concern by enabling multiple healthcare institutions to collaboratively learn from decentralized data without sharing it. FL’s scope in healthcare covers areas such as disease prediction, treatment customization, and clinical trial research. However, implementing FL poses challenges, including model convergence in non-IID (independent and identically distributed) data environments, communication overhead, and managing multi-institutional collaborations. A systematic review of FL in healthcare is necessary to evaluate how effectively FL can provide privacy while maintaining the integrity and usability of medical data analysis. In this study, we analyze existing literature on FL applications in healthcare. We explore the current state of model security practices, identify prevalent challenges, and discuss practical applications and their implications. Additionally, the review highlights promising future research directions to refine FL implementations, enhance data security protocols, and expand FL’s use to broader healthcare applications, which will benefit future researchers and practitioners. © 2024, CC BY.

关键词： Clinical research

Assessing the Performance of the TOFPET2c ASIC at X-ray Energies for Time-of-Flight Scatter Rejection Applications

学校读者我要写书评

暂无评论

Assessing the Performance of the TOFPET2c ASIC at X-ray Ener...

IEEE Symposium on Nuclear Science (NSS/MIC)

作者： L.-D. Gaulin V. Nadig J. Rossignol F. Gagnon K. Herweg V. Schulz S. Gundacker R. Fontaine Departement of electrical and computer engineering Université de Sherbrooke Sherbrooke Québec Canada Department of Physics of Molecular Imaging Systems RWTH Aachen University Aachen North Rhine-Westphalia Germany

ISBN: (数字)9798350388152

ISBN: (纸本)9798350388169

The addition of a time-of-flight (ToF) measurements to radiography and computed tomography (CT) opens the door to an anti-scatter grid-free approach to scatter rejection in imaging systems, potentially increasing system sensitivity and image quality. Previously developed hardware limited to a few channels showed that the ToF scatter rejection is possible, but lacked in scale and density. A medium-scale ToF scatter rejection detection module was developed, allowing for the evaluation of off-the-shelf ToF ASICs, as well as the future development of dedicated devices. The TOFPET2c ASIC designed by PETSys showed good potential for the first detector, offering high integration and pixel count. The evaluation of the TOFPET2c ASIC at X-ray energies (20 keV to 140 keV) showed a 414 ps FWHM detector timing resolution. Such a value is slightly over the 300 ps FWHM limit sufficient to reach scatter-rejection similar to anti-scatter grids and requires further investigation. An energy resolution bellow 50% was also reached for all desired energies.

关键词： Radiography Image quality Semiconductor device measurement Microwave integrated circuits Sensitivity Semiconductor detectors Computed tomography

Three-Dimensional Nuclei Synthesis for Fluorescence Microscopy Image Analysis

学校读者我要写书评

暂无评论

Three-Dimensional Nuclei Synthesis for Fluorescence Microsco...

IEEE International Symposium on Biomedical Imaging

作者： Liming Wu Alain Chen Paul Salama Kenneth W. Dunn Edward J. Delp Video and Image Processing Laboratory School of Electrical and Computer Engineering Purdue University West Lafayette Indiana USA Department of Electrical and Computer Engineering Indiana University-Purdue University Indianapolis Indiana USA Division of Nephrology School of Medicine Indiana University Indianapolis Indiana USA

Three-dimensional tissue cytometry is an important technique for quantitative analysis of cell structures in large fluorescence microscopy volumes. Accurate nuclei detection and segmentation is an important step for 3D tissue cytometry. Deep learning methods have shown promising results for nuclei detection and segmentation. However, manually annotating ground truth for training deep learning methods is labor-intensive and not practical for large 3D volumes. In this paper, we propose a 3D nuclei synthesis method, known as 3DSpCycleGAN, for generating 3D ground truth volumes along with corresponding synthetic microscopy volumes. Experimental results using fluorescence microscopy volumes demonstrate that our method generates more realistic 3D volumes when evaluated both visually and quantitatively than previously reported. We also show that using the synthetic volumes generated by 3DSpCycleGAN as training data improves segmentation accuracy for deep learning segmentation techniques.

关键词：

Cooperative Control for Motor Drive System Integrated of Multiple High-Speed Motors and Magnetic gear

学校读者我要写书评

暂无评论

Cooperative Control for Motor Drive System Integrated of Mul...

IEEE Energy Conversion Congress and Exposition, ECCE

作者： Kohei Aiso Kan Akatsu Yasuaki Aoyama Department of Electrical Engineering Shibaura Institute of Technology Tokyo Japan Department of Faculty of Engineering Division of Intelligent Systems Engineering Yokohama National University Yokohama Japan Research & Development Group Center for Technology Innovation – Materials Hitachi Ltd. Ibaraki Japan

A novel motor drive system integrated of multiple high speed motors and magnetic gear which called as Magnetic Multiple Spur Gear (MMSG) efficiently converts the total output power of high speed motors by using MMSG in non-contact, that is effective to realize the downsizing, high power density, and high efficiency for motor systems. However, it causes the torque unbalance between multiple motors by the effects of the difference of individual motor parameters such as inductance and winding resistance, and the detection error of magnetic poles position that causes unbalanced load between motors, resulting in lower efficiency and out-of-synchronization of magnetic gear. Moreover, the simple and inexpensive control system is required for driving multiple motors. In this paper, the behavior of MMSG when torque unbalance occurs is clarified, the cooperative control for torque unbalance between multiple motors is proposed.

关键词：

The Mechanism of Orientation Detection Based on Local Orientation-Selective Neuron 6

学校读者我要写书评

暂无评论

The Mechanism of Orientation Detection Based on Local Orient...

6th International Conference on Computational Intelligence and Applications, ICCIA 2021

作者： Li, Bin Todo, Yuki Tang, Zheng Kanazawa University Faculty of Electrical and Computer Engineering Kanazawa-shi9201192 Japan University of Toyama Information Systems Department of Intelligence Toyama Japan

ISBN: (纸本)9781665439336

Orientation detection is an essential function of the visual system. It is a basic behavior among creatures in nature and can directly affect animals' behavioral decisions. Previously, Hubel and Wiesel studied cats' and monkeys' striate cortex and observed that some cells showed orientation selectivity in their response to one object. However, the mechanism of orientation selectivity, especially the global orientation detection of an object, is still an open problem. To Figure out this issue, we propose a new orientation detection mechanism based on local orientation-selective neurons to solve two-dimensional detection tasks. We assume that there are neurons only responsible for orientation detection in the brain, and each neuron is responsible only for detecting a local specific orientation. We use the McCulloch-Pitts neuron model to realize such neurons. The global orientation is obtained according to the number of orientation-selective neurons be activated. We performed computer simulations based on this mechanism, and the result shows that this mechanism works well for global orientation detection. This novel mechanism may lead to a brand-new pattern recognition method and promote the visual nervous system's development. © 2021 IEEE.

关键词： Neurons

Multi-Resolution Online Deterministic Annealing: A Hierarchical and Progressive Learning Architecture

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Mavridis, Christos N. Baras, John S. Department of Electrical and Computer Engineering Institute for Systems Research University of Maryland College Park United States

Hierarchical learning algorithms that gradually approximate a solution to a data-driven optimization problem are essential to decision-making systems, especially under limitations on time and computational resources. In this study, we introduce a general-purpose hierarchical learning architecture that is based on the progressive partitioning of a possibly multi-resolution data space. The optimal partition is gradually approximated by solving a sequence of optimization sub-problems online, using gradient-free stochastic approximation updates. As a consequence, a function approximation problem can be defined within each subset of the partition and solved using the theory of two-timescale stochastic approximation. This simulates an annealing process and defines a robust and interpretable heuristic method to gradually increase the complexity of the learning architecture in a task-agnostic manner, giving emphasis to regions of the data space that are considered more important according to a predefined criterion. Finally, by imposing a tree structure in the progression of the partitions, we provide a means to incorporate potential multi-resolution structure of the data space into this approach, significantly reducing its complexity, while introducing hierarchical variable-rate feature extraction properties similar to certain classes of deep learning architectures. Asymptotic convergence analysis and experimental results are provided for supervised and unsupervised learning problems. © 2022, CC BY.

关键词： Learning algorithms