Beihang University

About: Beihang University is a education organization based out in Beijing, China. It is known for research contribution in the topics: Computer science & Control theory. The organization has 67002 authors who have published 73507 publications receiving 975691 citations. The organization is also known as: Beijing University of Aeronautics and Astronautics.

Correlation Congruence for Knowledge Distillation

Abstract: Most teacher-student frameworks based on knowledge distillation (KD) depend on a strong congruent constraint on instance level. However, they usually ignore the correlation between multiple instances, which is also valuable for knowledge transfer. In this work, we propose a new framework named correlation congruence for knowledge distillation (CCKD), which transfers not only the instance-level information but also the correlation between instances. Furthermore, a generalized kernel method based on Taylor series expansion is proposed to better capture the correlation between instances. Empirical experiments and ablation studies on image classification tasks (including CIFAR-100, ImageNet-1K) and metric learning tasks (including ReID and Face Recognition) show that the proposed CCKD substantially outperforms the original KD and other SOTA KD-based methods. The CCKD can be easily deployed in the majority of the teacher-student framework such as KD and hint-based learning methods.

Remaining Useful Life Prediction for Lithium-Ion Battery: A Deep Learning Approach

Abstract: Accurate prediction of remaining useful life (RUL) of lithium-ion battery plays an increasingly crucial role in the intelligent battery health management systems. The advances in deep learning introduce new data-driven approaches to this problem. This paper proposes an integrated deep learning approach for RUL prediction of lithium-ion battery by integrating autoencoder with deep neural network (DNN). First, we present a multi-dimensional feature extraction method with autoencoder model to represent battery health degradation. Then, the RUL prediction model-based DNN is trained for multi-battery remaining cycle life estimation. The proposed approach is applied to the real data set of lithium-ion battery cycle life from NASA, and the experiment results show that the proposed approach can improve the accuracy of RUL prediction.

Two-Dimensional Amorphous TiO2 Nanosheets Enabling High-Efficiency Photoinduced Charge Transfer for Excellent SERS Activity.

Abstract: Substrate–molecule vibronic coupling enhancement, especially the efficient photoinduced charge transfer (PICT), is pivotal to the performance of nonmetal surface-enhanced Raman scattering (SERS) te...

Porous Core–Shell Fe3C Embedded N-doped Carbon Nanofibers as an Effective Electrocatalysts for Oxygen Reduction Reaction

Abstract: The development of nonprecious-metal-based electrocatalysts with high oxygen reduction reaction (ORR) activity, low cost, and good durability in both alkaline and acidic media is very important for application of full cells. Herein, we developed a facile and economical strategy to obtain porous core–shell Fe3C embedded nitrogen-doped carbon nanofibers (Fe3C@NCNF-X, where X denotes pyrolysis temperature) by electrospinning of polyvinylidene fluoride (PVDF) and FeCl3 mixture, chemical vapor phase polymerization of pyrrole, and followed by pyrolysis of composite nanofibers at high temperatures. Note that the FeCl3 and polypyrrole acts as precursor for Fe3C core and N-doped carbon shell, respectively. Moreover, PVDF not only plays a role as carbon resources, but also provides porous structures due to hydrogen fluoride exposure originated from thermal decomposition of PVDF. The resultant Fe3C@NCNF-X catalysts, particularly Fe3C@NCNF-900, showed efficient electrocatalytic performance for ORR in both alkaline an...

Support Vector Machine Soft Margin Classifiers: Error Analysis

Abstract: The purpose of this paper is to provide a PAC error analysis for the q-norm soft margin classifier, a support vector machine classification algorithm. It consists of two parts: regularization error and sample error. While many techniques are available for treating the sample error, much less is known for the regularization error and the corresponding approximation error for reproducing kernel Hilbert spaces. We are mainly concerned about the regularization error. It is estimated for general distributions by a K-functional in weighted Lq spaces. For weakly separable distributions (i.e., the margin may be zero) satisfactory convergence rates are provided by means of separating functions. A projection operator is introduced, which leads to better sample error estimates especially for small complexity kernels. The misclassification error is bounded by the V-risk associated with a general class of loss functions V. The difficulty of bounding the offset is overcome. Polynomial kernels and Gaussian kernels are used to demonstrate the main results. The choice of the regularization parameter plays an important role in our analysis.