In this study, we focus on developing efficient calibration methods via Bayesian decision-making for the family of compartmental epidemiological models. The existing calibration methods usually assume that the compart...
详细信息
We model an all-photonic quantum simulator that consists of a solid-state coupled-cavity array with integrated ensembles of color centers. We quantify conditions for polariton creation that overcome the setbacks of mo...
详细信息
As a widely used loss function in deep face recognition, the softmax loss cannot guarantee that the minimum positive sample-to-class similarity is larger than the maximum negative sample-to-class similarity. As a resu...
As a widely used loss function in deep face recognition, the softmax loss cannot guarantee that the minimum positive sample-to-class similarity is larger than the maximum negative sample-to-class similarity. As a result, no unified threshold is available to separate positive sample-to-class pairs from negative sample-to-class pairs. To bridge this gap, we design a UCE (Unified Cross-Entropy) loss for face recognition model training, which is built on the vital constraint that all the positive sample-to-class similarities shall be larger than the negative ones. Our UCE loss can be integrated with margins for a further performance boost. The face recognition model trained with the proposed UCE loss, UniFace, was intensively evaluated using a number of popular public datasets like MFR, IJB-C, LFW, CFP-FP, AgeDB, and MegaFace. Experimental results show that our approach outperforms SOTA methods like SphereFace, CosFace, ArcFace, Partial FC, etc. Especially, till the submission of this work (Mar. 8, 2023), the proposed UniFace achieves the highest TAR@MR-All on the academic track of the MFR-ongoing challenge. $\color{Blue}{\mathbf{Code}}$ is publicly available.
Low-disorder two-dimensional electron systems in the presence of a strong, perpendicular magnetic field terminate at very small Landau level filling factors in a Wigner crystal (WC), where the electrons form an ordere...
详细信息
Quantum radar uses the strong correlations of entangled photons and has a great potential to improve the detection sensitivity. Recently, many research results have been achieved in theoretical analysis and experiment...
详细信息
Low temperature fuel cells (LTFCs) present promising technical solutions for enabling broad applications of green Hydrogen energy. However, the efficiencies and durability of LTFCs are limited by the dynamic water evo...
详细信息
ISBN:
(数字)9798350371901
ISBN:
(纸本)9798350371918
Low temperature fuel cells (LTFCs) present promising technical solutions for enabling broad applications of green Hydrogen energy. However, the efficiencies and durability of LTFCs are limited by the dynamic water evolution and transport, especially when they are operated at high current densities. Developing water monitoring techniques therefore presents a crucial strategy for optimizing the designs and operating schemes of LTFCs, and eventually overcoming such water management issues. This work explores the possibilities of using ultrasonic techniques to investigate the water contents in the gas diffusion layers (GDL), as one of the key components of the membrane electrode assembly in LTFCs. For this purpose, we numerically and experimentally investigated the propagation of longitudinal mode ultrasonic waves in GDLs at different water saturation levels. The numerical simulations and experimental investigations both reveal the changes in the speed of sound and acoustic attenuation coefficient with respect to the water saturation level of the GDL. These physical effects open new horizons for developing simplified and cost-effective ultrasound based water monitoring systems towards applications in operating LTFCs.
THE USE OF KNOWLEDGE GRAPH IN NATURAL SCIENCE Knowledge graph is a field of Artificial Intelligence(AI)that aims to represent knowledge in the form of graphs,consisting of nodes and edges which represent entities and ...
详细信息
THE USE OF KNOWLEDGE GRAPH IN NATURAL SCIENCE Knowledge graph is a field of Artificial Intelligence(AI)that aims to represent knowledge in the form of graphs,consisting of nodes and edges which represent entities and relationships between nodes respectively(Aidan et al.,2022).Although the knowledge graph was popularized recently due to use of this idea in Google’s search engine in 2012(Amit,2012),its root can be traced back to the emergence of the Semantic Web as well as earlier works in ontology(Aggarwal,2021).
This paper focuses on the DC-bus control of asymmetrical Multilevel inverter family featuring a Buck-Boost converter to boost the input voltage. To investigate the dynamic performance of the system a thorough analysis...
详细信息
ISBN:
(数字)9789075815399
ISBN:
(纸本)9781665487009
This paper focuses on the DC-bus control of asymmetrical Multilevel inverter family featuring a Buck-Boost converter to boost the input voltage. To investigate the dynamic performance of the system a thorough analysis is presented on the DC-bus dynamic behavior. It is shown that with higher system bandwidth the input capacitance requirements and the peak current through the boosting inductor are increased, compromising the power density and the efficiency of the whole DC/AC converter. To improve the transient performance of the system without increasing the volume of the passive components, a control scheme with a feedforward current estimator term is proposed. Finally, the correctness of the theoretical analysis is validated via experimental probing on a laboratory prototype.
The functional processes and the upgrading strategies in a company environment depend not only on the firm’s internal and external sourcing strategies but also on international experience and interactions developed o...
Cooperative co-evolutionary algorithms (CCEAs) have witnessed giant success in solving large-scale optimization problems (LSOPs). However, most existing CCEAs use low-dimensional EAs to optimize the decomposed sub-pro...
详细信息
暂无评论