How to effectively compress mechanical signals so that they can support remote and real-time health monitoring is a hot issue in the context of intelligent manufacturing. Therefore, this paper presents a novel mechani...
详细信息
The behaviors exhibited by stored grain pests, including feeding, respiration, excretion, and reproduction, engender multifarious threats to grain storage. Consequently, it is imperative to diligently undertake the ta...
详细信息
This paper provides a novel estimation procedure for determining the number of high-dimensional complex-valued signals embedded in Gaussian white noise. Initially, this problem is formulated as a sequence of nested hy...
ISBN:
(数字)9781837240982
This paper provides a novel estimation procedure for determining the number of high-dimensional complex-valued signals embedded in Gaussian white noise. Initially, this problem is formulated as a sequence of nested hypothesis tests. To enhance the decision performance, we introduce new test statistics based on the ratio of the fourth- and square second-order moments of the population covariance eigenvalues. This enables us to leverage higher-order moments information in the observed data, distinguishing it from conventional tests that rely solely on the first and second order moments. In the context of high-dimensional, complex-valued data and within a sub-sphericity testing framework, unbiased and consistent estimates are derived for these two higher-order moments. Subsequently, we conduct an asymptotic analysis to incoporate these estimates into the construction of the test statistics and the design of our proposed estimation scheme for determining the number of sources. Finally, numerical examples are presented to show its effectiveness and superiority over several classical estimation methods.
The Back-n white neutron source(known as Back-n)is based on back-streaming neutrons from the spallation target at the China Spallation Neutron Source(CSNS).With its excellent beam properties,e.g.,a neutron flux of app...
详细信息
The Back-n white neutron source(known as Back-n)is based on back-streaming neutrons from the spallation target at the China Spallation Neutron Source(CSNS).With its excellent beam properties,e.g.,a neutron flux of approximately 1.8×107 n/cm2/s at 55 m from the spallation target,energy range spanning from 0.5 eV to 200 MeV,and time-of-flight resolution of a few per thousand,along with the equipped physical spectrometers,Back-n is considered to be among the best facilities in the world for carrying out nuclear data *** its completion and commencement of operation in May 2018,five types of cross-section measurements concerning neutron capture cross-sections,fission cross-sections,total cross-sections,light charged particle emissions,in-beam gamma spectra,and more than forty nuclides have been *** article presents an overview of the experimental setup and result analysis on the neutron-induced cross-section measurements and gamma spectroscopy at Back-n in the initial years.
In the field of binocular stereo matching, remarkable progress has been made by iterative methods like RAFT-Stereo and CREStereo. However, most of these methods lose information during the iterative process, making it...
In the field of binocular stereo matching, remarkable progress has been made by iterative methods like RAFT-Stereo and CREStereo. However, most of these methods lose information during the iterative process, making it difficult to generate more detailed difference maps that take full advantage of high-frequency information. We propose the Decouple module to alleviate the problem of data coupling and allow features containing subtle details to transfer across the iterations which proves to alleviate the problem significantly in the ablations. To further capture high-frequency details, we propose a Normalization Refinement module that unifies the disparities as a proportion of the disparities over the width of the image, which address the problem of module failure in cross-domain scenarios. Further, with the above improvements, the ResNet-like feature extractor that has not been changed for years becomes a bottleneck. Towards this end, we proposed a multi-scale and multi-stage feature extractor that introduces the channel-wise self-attention mechanism which greatly addresses this bottleneck. Our method (DLNR) ranks 1st on the Middlebury leaderboard, significantly outperforming the next best method by 13.04%. Our method also achieves SOTA performance on the KITTI-2015 benchmark for D1-fg. Code and demos are available at: https://***/David-Zhao-1997/High-frequency-Stereo-Matching-Network.
We address the problem of learning new classes for semantic segmentation models from few examples, which is challenging because of the following two reasons. Firstly, it is difficult to learn from limited novel data t...
We address the problem of learning new classes for semantic segmentation models from few examples, which is challenging because of the following two reasons. Firstly, it is difficult to learn from limited novel data to capture the underlying class distribution. Secondly, it is challenging to retain knowledge for existing classes and to avoid catastrophic forgetting. For learning from limited data, we propose a pseudo-labeling strategy to augment the few-shot training annotations in order to learn novel classes more effectively. Given only one or a few images labeled with the novel classes and a much larger set of unlabeled images, we transfer the knowledge from labeled images to unlabeled images with a coarse-to-fine pseudo-labeling approach in two steps. Specifically, we first match each labeled image to its nearest neighbors in the unlabeled image set at the scene level, in order to obtain images with a similar scene layout. This is followed by obtaining pseudo-labels within this neighborhood by applying classifiers learned on the few-shot annotations. In addition, we use knowledge distillation on both labeled and unlabeled data to retain knowledge on existing classes. We integrate the above steps into a single convolutional neural network with a unified learning objective. Extensive experiments on the Cityscapes and KITTI datasets validate the efficacy of the proposed approach in the self-driving domain. Code is available from https://***/ChasonJiang/FSCILSS.
The angle-differential cross sections of neutron-induced deuteron production from carbon were measured at six neutron energies from 25 to 52 MeV relative to those of n-p elastic scattering at the China Spallation Neut...
详细信息
The angle-differential cross sections of neutron-induced deuteron production from carbon were measured at six neutron energies from 25 to 52 MeV relative to those of n-p elastic scattering at the China Spallation Neutron Source(CSNS)Back-n white neutron *** employing theΔE-E telescopes of the Light-charged Particle Detector Array(LPDA)system at 15.1°to 55.0°in the laboratory system,ratios of the angle-differential cross sections of the ^(12)C(n,xd)reactions to those of the n-p scattering were measured,and then,the angle-differential cross sections of the ^(12)C(n,xd)reactions were obtained using the angle-differential cross sections of the n-p elastic scattering from the JENDL-4.0/HE-2015 library as the *** obtained results are compared with data from previous measurements,all of which are based on mono-energic neutrons,the evaluated data from the JENDL-4.0/HE-2015 library and the ENDF-B/VIII.0 library,and those from theoretical calculations based on INCA code and Talys-1.9 *** the first white-neutron-source-based systematic measurement of the angle-differential cross sections of neutron-induced deuteron production reactions on carbon in several tens of MeV,the present work can provide a reference to the data library considering the lack of experimental data.
Neural Style Transfer (NST) exerted algorithms to generate animating images in computer vision for decades. Convolutional Neural Network (CNN) applied to image contents and styles in NST has improved feature extractio...
详细信息
This paper explores a novel multi-modal alternating learning paradigm pursuing a reconciliation between the exploitation of uni-modal features and the exploration of cross-modal interactions. This is motivated by the ...
This paper explores a novel multi-modal alternating learning paradigm pursuing a reconciliation between the exploitation of uni-modal features and the exploration of cross-modal interactions. This is motivated by the fact that current paradigms of multi-modal learning tend to explore multi-modal features simultaneously. The resulting gradient prohibits further exploitation of the features in the weak modality, leading to modality competition, where the dominant modality overpowers the learning process. To address this issue, we study the modality-alternating learning paradigm to achieve reconcilement. Specifically, we propose a new method called ReconBoost to update a fixed modality each time. Herein, the learning objective is dynamically adjusted with a reconcilement regularization against competition with the historical models. By choosing a KL-based reconcilement, we show that the proposed method resembles Friedman's Gradient-Boosting (GB) algorithm, where the updated learner can correct errors made by others and help enhance the overall performance. The major difference with the classic GB is that we only preserve the newest model for each modality to avoid overfitting caused by ensembling strong learners. Furthermore, we propose a memory consolidation scheme and a global rectification scheme to make this strategy more effective. Experiments over six multi-modal benchmarks speak to the efficacy of the method. We release the code at https://***/huacong/ReconBoost.
In order to facilitate government departments to assess security risks and prevent infiltration, it's necessary to recognize the IoT device from open data by the method of Named entity recognition (NER). In this s...
详细信息
暂无评论