Interest in the two-dimensional (2D) semiconducting transition metal dichalcogenides (TMDs) continues to intensify, driven by their suitable band gaps to supplant silicon as next-generation semiconductor materials. Am...
详细信息
Interest in the two-dimensional (2D) semiconducting transition metal dichalcogenides (TMDs) continues to intensify, driven by their suitable band gaps to supplant silicon as next-generation semiconductor materials. Among various TMDs, tungsten diselenide (WSe2) is renowned for its superior electrical properties in carrier density and mobility under ambient conditions. Despite its notable attributes, the behavior of monolayer WSe2 in the electron-doped regime under cryogenic conditions remains largely uncharted, particularly concerning its magnetotransport properties. In this study, we reveal the transport mechanisms of monolayer WSe2 from high temperatures down to the cryogenic regime. As evident by Efros–Shklovskii variable-range hopping (E-S VRH) in the cryogenic regime, strong Coulomb interactions arise between electrons. Above 8 K, an uncommon nonsaturated quadratic large magnetoresistance (MR) can be explained by the wave-function shrinkage model, which is consistent with the E-S VRH transport mechanism. Notably, the nonsaturated quadratic large MR shows a magnitude up to 1740% at 13 T. These findings underscore the potential applications for monolayer WSe2 in cryogenic field-effect devices, magnetic sensors, and memory devices and mark a significant advance in magnetotransport research.
In the evolving landscape of ECG signal analysis, the challenge of limited transparency in machine learning models remains a significant barrier to their effective integration into clinical practice. This study addres...
详细信息
Semi-supervised learning (SSL) essentially pursues class boundary exploration with less dependence on human annotations. Although typical attempts focus on ameliorating the inevitable error-prone pseudo-labeling, we t...
Semi-supervised learning (SSL) essentially pursues class boundary exploration with less dependence on human annotations. Although typical attempts focus on ameliorating the inevitable error-prone pseudo-labeling, we think differently and resort to exhausting informative semantics from multiple probably correct candidate labels. In this paper, we introduce Fuzzy Positive Learning (FPL) for accurate SSL semantic segmentation in a plug-and-play fashion, targeting adaptively encouraging fuzzy positive predictions and suppressing highly-probable negatives. Being conceptually simple yet practically effective, FPL can remarkably alleviate interference from wrong pseudo labels and progressively achieve clear pixel-level semantic discrimination. Concretely, our FPL approach consists of two main components, including fuzzy positive assignment (FPA) to provide an adaptive number of labels for each pixel and fuzzy positive regularization (FPR) to restrict the predictions of fuzzy positive categories to be larger than the rest under different perturbations. Theoretical analysis and extensive experiments on Cityscapes and VOC 2012 with consistent performance gain justify the superiority of our approach. Codes are provided in https://***/qpc1611094/FPL.
An e-Commerce company has been using an Enterprise Resource Planning (ERP) system for several years, but is still constrained in its implementation, this is reflected in the number of issue/change request tickets subm...
详细信息
Summary: Named entity recognition (NER) is a fundamental part of extracting information from documents in biomedical applications. A notable advantage of NER is its consistency in extracting biomedical entities in a d...
详细信息
The electromagnetic (EM) interaction between two anisotropic nonlinear metasurfaces, under arbitrary rotations of their optical axes, is rigorously studied. Multistability conditions with respect to rotation angles be...
详细信息
Recent advances in robust semi-supervised learning (SSL) typically filter out-of-distribution (OOD) information at the sample level. We argue that an overlooked problem of robust SSL is its corrupted information on se...
Recent advances in robust semi-supervised learning (SSL) typically filter out-of-distribution (OOD) information at the sample level. We argue that an overlooked problem of robust SSL is its corrupted information on semantic level, practically limiting the development of the field. In this paper, we take an initial step to explore and propose a unified framework termed OOD Semantic Pruning (OSP), which aims at pruning OOD semantics out from in-distribution (ID) features. Specifically, (i) we propose an aliasing OOD matching module to pair each ID sample with an OOD sample with semantic overlap. (ii) We design a soft orthogonality regularization, which first transforms each ID feature by suppressing its semantic component that is collinear with paired OOD sample. It then forces the predictions before and after soft orthogonality decomposition to be consistent. Being practically simple, our method shows a strong performance in OOD detection and ID classification on challenging benchmarks. In particular, OSP surpasses the previous state-of-the-art by 13.7% on accuracy for ID classification and 5.9% on AUROC for OOD detection on TinyImageNet dataset. The source codes are publicly available at https://***/rain305f/OSP.
This article describes the implementation of an advanced fiber-optic intensity sensor for the comprehensive detection of axles and bogies of rail vehicles in tram transport. The sensor uses the principle of optical fi...
详细信息
Methods for harnessing vibrational states are desired for phonon-based technologies. We realized ultrastrong coupling of two phonon modes in perovskite materials induced by ultrastrong coupling with a common photonic ...
详细信息
The proliferation of digital camera technology has provided many digital images for object detection and counting in public places. This paper presents an experiment on the realtime vehicle counting method to be imple...
详细信息
ISBN:
(纸本)9781665499705
The proliferation of digital camera technology has provided many digital images for object detection and counting in public places. This paper presents an experiment on the realtime vehicle counting method to be implemented in a single board computer. The proposed design's objective is to detect and count vehicles entering and exiting at entrances using haar cascade classifier algorithms and other automatic open and close door systems controlled by a single board computer (SBC). The proposed system was tested using raspberry pi 4 model B with a 4GB RAM platform. The empiric results using 3-minute video as input with a total of 113 passing vehicles the proposed method achieve of 91.2 % of average accuracy.
暂无评论