In this paper, we study the global well-posedness of the 2D compressible NavierStokes equations with large initial data and vacuum. It is proved that if the shear viscosity<i></i> is a positive constant and the bulk viscosity <i></i> is the power function of the density, that is, <i></i>(<i></i>) = <i></i> ^<i></i> with <i></i> &gt; 3, then the 2D compressible NavierStokes equations with the periodic boundary conditions on the torus T 2 admit a unique global classical solution (<i>, u</i>) which may contain vacuums in an open set of T 2 . Note that the initial data can be arbitrarily large to contain vacuum states.

The purpose of this research monograph is to survey some recent developments in the analysis of shock reflection-diffraction, to present our original mathematical proofs of von Neumann’s conjectures for potential flow, to collect most of the related results and new techniques in the analysis of partial differential equations achieved in the last decades, and to discuss a set of fundamental open problems relevant to the directions of future research in this and related areas.

For a fixed geometric configuration, hydrodynamic instabilities and bifurcation processes of laminar isothermal planar opposed jet flows with symmetric and slightly asymmetric inlet boundary conditions are investigated numerically using a high-resolution approach based on spectral element method. In current configuration, when inlet boundary conditions are symmetric, in the range of the Reynolds number considered (Re <= 200), multiple new symmetry-breaking bifurcations are observed and four new flow patterns are identified. Their hydrodynamic characteristics are analyzed, in particular their symmetries. In addition, the case that inlet boundary conditions are slightly asymmetric is investigated. It is found that bifurcation processes are extremely sensitive to this small symmetry-breaking imperfection and much different from those in the symmetric case. Furthermore, model equations are constructed by symmetry consideration to explain the numerical results based on hydrodynamic equations.

Training process is crucial for the deployment of the network in applications which have two strict requirements on both accuracy and robustness. However, most existing approaches are in a dilemma, i.e. model accuracy and robustness forming an embarrassing tradeoff – the improvement of one leads to the drop of the other. The challenge remains for as we try to improve the accuracy and robustness simultaneously. In this paper, we propose a novel training method via introducing the auxiliary classifiers for training on corrupted samples, while the clean samples are normally trained with the primary classifier. In the training stage, a novel distillation method named input-aware self distillation is proposed to facilitate the primary classifier to learn the robust information from auxiliary classifiers. Along with it, a new normalization method - selective batch normal- ization is proposed to prevent the model from the negative influence of corrupted images. At the end of the training period, a L2-norm penalty is applied to the weights of primary and auxiliary classifiers such that their weights are asymptotically identical. In the stage of inference, only the primary classifier is used and thus no extra computation and storage are needed. Extensive experiments on CIFAR10, CIFAR100 and ImageNet show that noticeable improvements on both accuracy and robustness can be observed by the proposed auxiliary training. On average, auxiliary training achieves 2.21% accuracy and 21.64% robustness (measured by corruption error) improvements over traditional training methods on CIFAR100. Codes have been released on github.

After Bershadsky-Cecotti-Ooguri-Vafa, we introduce an invariant of Calabi-Yau threefolds, which we call the BCOV invariant and which we obtain using analytic torsion. We give an explicit formula for the BCOV invariant as a function on the compactified moduli space, when it is isomorphic to a projective line. As a corollary, we prove the formula for the BCOV invariant of quintic mirror threefolds conjectured by Bershadsky-Cecotti-Ooguri-Vafa.