• Multi-shifted linear systems with non-Hermitian coefficient matrices arise in numerical solutions of time-dependent partial/fractional differential equations (PDEs/FDEs), in control theory, PageRank problems, and other research fields. We derive efficient variants of the restarted Changing Minimal Residual method based on the cost-effective Hessenberg procedure (CMRH) for this problem class. Then, we introduce a flexible variant of the algorithm that allows to use variable preconditioning at each iteration to further accelerate the convergence of shifted CMRH. We analyse the performance of the new class of methods in the numerical solution of PDEs and FDEs, also against other multi-shifted Krylov subspace methods.
  • We show that in the space of all convex billiard boundaries, the set of boundaries with rational caustics is dense. More precisely, the set of billiard boundaries with caustics of rotation number $1/q$ is polynomial sense in the smooth case, and exponentially dense in the analytic case.
  • Herbig Ae/Be (HAeBe) stars have been classified into Group I or Group II, which were initially thought to be flared and flat disks, respectively. Several Group I sources have been shown to have large gaps, suggesting ongoing planet formation, while no large gaps have been found in the disks of Group II sources. We analyzed the disk around the Group II source, HD 142666, using irradiated accretion disk modeling of the broad-band spectral energy distribution along with the 1.3 millimeter spatial brightness distribution traced by Atacama Large Millimeter and Submillimeter Array (ALMA) observations. Our model reproduces the available data, predicting a high degree of dust settling in the disk, which is consistent with the Group II classification of HD 142666. In addition, the observed visibilities and synthesized image could only be reproduced when including a depletion of large grains out to ~16 au in our disk model, although the ALMA observations did not have enough angular resolution to fully resolve the inner parts of the disk. These results may suggest that some disks around Group II HAeBe stars have cavities of large grains as well. Further ALMA observations of Group II sources are needed to discern how commonly cavities occur in this class of objects, as well as to reveal their possible origins.
  • We show that the paratingent cone of the Aubry set of the Tonelli Hamiltonian is contained in a cone bounded by the Green bundles. Our result improves the earlier result of M.-C. Arnaud on tangent cones of the Aubry sets.
  • In the present paper we prove a strong form of Arnold diffusion. Let $\mathbb{T}^2$ be the two torus and $B^2$ be the unit ball around the origin in $\mathbb{R}^2$. Fix $\rho>0$. Our main result says that for a "generic" time-periodic perturbation of an integrable system of two degrees of freedom \[ H_0(p)+\epsilon H_1(\theta,p,t),\quad \ \theta\in \mathbb{T}^2,\ p\in B^2,\ t\in \mathbb{T}, \] with a strictly convex $H_0$, there exists a $\rho$-dense orbit $(\theta_{\epsilon},p_{\epsilon},t)(t)$ in $\mathbb{T}^2 \times B^2 \times \mathbb{T}$, namely, a $\rho$-neighborhood of the orbit contains $\mathbb{T}^2 \times B^2 \times \mathbb{T}$. Our proof is a combination of geometric and variational methods. The fundamental elements of the construction are usage of crumpled normally hyperbolic invariant cylinders from \cite{BKZ}, flower and simple normally hyperbolic invariant manifolds from as well as their kissing property at a strong double resonance. This allows us to build a "connected" net of $3$-dimensional normally hyperbolic invariant manifolds. To construct diffusing orbits along this net we employ a version of Mather variational method \cite{Ma2} proposed by Bernard in \cite{Be}. This version is equipped with weak KAM theory \cite{Fa}.
  • At a time when ALMA produces spectacular high resolution images of gas and dust in circumstellar disks, the next observational frontier in our understanding of planet formation and the chemistry of planet-forming material may be found in the mid- to far-infrared wavelength range. A large, actively cooled far-infrared telescope in space will offer enormous spectroscopic sensitivity improvements of 3-4 orders of magnitude, making it possible to uniquely survey certain fundamental properties of planet formation. Specifically, the Origins Space Telescope (OST), a NASA flagship concept to be submitted to the 2020 decadal survey, will provide a platform that allows complete surveys of warm and cold water around young stars of all masses and across all evolutionary stages, and to measure their total planet-forming gas mass using the ground-state line of HD. While this white paper is formulated in the context of the NASA Origins Space Telescope concept, it can be applied in general to inform any future space-based, cold far-infrared observatory.
  • This paper describes a problem arising in sea exploration, where the aim is to schedule the expedition of a ship for collecting information about the resources on the seafloor. The aim is to collect data by probing on a set of carefully chosen locations, so that the information available is optimally enriched. This problem has similarities with the orienteering problem, where the aim is to plan a time-limited trip for visiting a set of vertices, collecting a prize at each of them, in such a way that the total value collected is maximum. In our problem, the score at each vertex is associated with an estimation of the level of the resource on the given surface, which is done by regression using Gaussian processes. Hence, there is a correlation among scores on the selected vertices; this is a first difference with respect to the standard orienteering problem. The second difference is the location of each vertex, which in our problem is a freely chosen point on a given surface.
  • Thermal conductivity of binary aqueous solutions of 1,2-ethanediol and 1,2-propanediol was measured using the transient hot wire method at temperature from 253.15 K to 373.15 K at atmospheric. Measurements were made for six compositions over the entire concentration range from 0 to 1 mole fraction of glycol, namely, 0.0, 0.2, 0.4, 0.6, 0.8, and 1.0 mole fraction of glycol. The uncertainty of the thermal conductivity, temperature, and concentration measurements are estimated to be better than 2%, 0.01K, 0.1%, respectively. The second-order Scheff\'e polynomial is used to correlate the temperature and composition dependence of the experimental thermal conductivity, which is found to be in good agreement with the experiment data from the present work and other reports.
  • The possible detection of C_{24}, a planar graphene, recently reported in several planetary nebulae by Garciaa-Hernandez et al. (2011, 2012) inspires us to explore whether and how much graphene could exist in the interstellar medium (ISM) and how it would reveal its presence through its ultraviolet (UV) extinction and infrared (IR) emission. In principle, interstellar graphene could arise from the photochemical processing of polycyclic aromatic hydrocarbon (PAH) molecules which are abundant in the ISM through a complete loss of their hydrogen atoms and/or from graphite which is thought to be a major dust species in the ISM through fragmentation caused by grain-grain collisional shattering. Both quantum-chemical computations and laboratory experiments have shown that the exciton-dominated electronic transitions in graphene cause a strong absorption band near 2755 Angstrom. We calculate the UV absorption of graphene and place an upper limit of ~5 ppm of C/H (i.e., ~1.9\% of the total interstellar C) on the interstellar graphene abundance. We also model the stochastic heating of graphene C_{24} excited by single starlight photons of the interstellar radiation field in the ISM and calculate its IR emission spectra. We also derive the abundance of graphene in the ISM to be <5 ppm of C/H by comparing the model emission spectra with that observed in the ISM.
  • Automatically predicting age group and gender from face images acquired in unconstrained conditions is an important and challenging task in many real-world applications. Nevertheless, the conventional methods with manually-designed features on in-the-wild benchmarks are unsatisfactory because of incompetency to tackle large variations in unconstrained images. This difficulty is alleviated to some degree through Convolutional Neural Networks (CNN) for its powerful feature representation. In this paper, we propose a new CNN based method for age group and gender estimation leveraging Residual Networks of Residual Networks (RoR), which exhibits better optimization ability for age group and gender classification than other CNN architectures.Moreover, two modest mechanisms based on observation of the characteristics of age group are presented to further improve the performance of age estimation.In order to further improve the performance and alleviate over-fitting problem, RoR model is pre-trained on ImageNet firstly, and then it is fune-tuned on the IMDB-WIKI-101 data set for further learning the features of face images, finally, it is used to fine-tune on Adience data set. Our experiments illustrate the effectiveness of RoR method for age and gender estimation in the wild, where it achieves better performance than other CNN methods. Finally, the RoR-152+IMDB-WIKI-101 with two mechanisms achieves new state-of-the-art results on Adience benchmark.
  • The Residual Networks of Residual Networks (RoR) exhibits excellent performance in the image classification task, but sharply increasing the number of feature map channels makes the characteristic information transmission incoherent, which losses a certain of information related to classification prediction, limiting the classification performance. In this paper, a Pyramidal RoR network model is proposed by analysing the performance characteristics of RoR and combining with the PyramidNet. Firstly, based on RoR, the Pyramidal RoR network model with channels gradually increasing is designed. Secondly, we analysed the effect of different residual block structures on performance, and chosen the residual block structure which best favoured the classification performance. Finally, we add an important principle to further optimize Pyramidal RoR networks, drop-path is used to avoid over-fitting and save training time. In this paper, image classification experiments were performed on CIFAR-10/100 and SVHN datasets, and we achieved the current lowest classification error rates were 2.96%, 16.40% and 1.59%, respectively. Experiments show that the Pyramidal RoR network optimization method can improve the network performance for different data sets and effectively suppress the gradient disappearance problem in DCNN training.
  • We effect thermodynamical formalism for the non-uniformly hyperbolic $C^\infty$ map of the two dimensional torus known as the Katok map. It is a slowdown of a linear Anosov map near the origin and it is a local (but not small) perturbation. We prove the existence of equilibrium measures for any continuous potential function and obtain uniqueness of equilibrium measures associated to the geometric $t$-potential $\varphi_t=-t\log |df|_{E^u(x)}|$ for any $t\in(t_0,\infty)$, $t\neq 1$ where $E^u(x)$ denotes the unstable direction. We show that $t_0$ tends to $-\infty$ as the size of the perturbation tends to zero. Finally, we establish exponential decay of correlations as well as the Central Limit Theorem for the equilibrium measures associated to $\varphi_t$ for all values of $t\in (t_0, 1)$.
  • The initial mass distribution in the solar nebula is a critical input to planet formation models that seek to reproduce today's Solar System. Traditionally, constraints on the gas mass distribution are derived from observations of the dust emission from disks, but this approach suffers from large uncertainties in grain growth and gas-to-dust ratio. On the other hand, previous observations of gas tracers only probe surface layers above the bulk mass reservoir. Here we present the first partially spatially resolved observations of the $^{13}$C$^{18}$O J=3-2 line emission in the closest protoplanetary disk, TW Hya, a gas tracer that probes the bulk mass distribution. Combining it with the C$^{18}$O J=3-2 emission and the previously detected HD J=1-0 flux, we directly constrain the mid-plane temperature and optical depths of gas and dust emission. We report a gas mass distribution of 13$^{+8}_{-5}\times$(R/20.5AU)$^{-0.9^{+0.4}_{-0.3}}$ g cm$^{-2}$ in the expected formation zone of gas and ice giants (5-21AU). We find the total gas/millimeter-sized dust mass ratio is 140 in this region, suggesting that at least 2.4M_earth of dust aggregates have grown to >centimeter sizes (and perhaps much larger). The radial distribution of gas mass is consistent with a self-similar viscous disk profile but much flatter than the posterior extrapolation of mass distribution in our own and extrasolar planetary systems.
  • Using a multi-layered printed circuit board, we propose a 3D architecture suitable for packaging supercon- ducting chips, especially chips that contain two-dimensional qubit arrays. In our proposed architecture, the center strips of the buried coplanar waveguides protrude from the surface of a dielectric layer as contacts. Since the contacts extend beyond the surface of the dielectric layer, chips can simply be flip-chip packaged with on-chip receptacles clinging to the contacts. Using this scheme, we packaged a multi-qubit chip and per- formed single-qubit and two-qubit quantum gate operations. The results indicate that this 3D architecture provides a promising scheme for scalable quantum computing.
  • We show that for a family of randomly kicked Hamiton-Jacobi equations on the torus, almost surely, the solution of an initial value problem converges exponentially fast to the unique stationary solution. Combined with the results in \cite{IK03} and \cite{KZ12}, this completes the program started in \cite{EKMS00} for the multi-dimensional setting.
  • We show that for a family of randomly kicked Hamilton-Jacobi equations, the unique global minimizer is hyperbolic, almost surely. Furthermore, we prove the unique forward and backward viscosity solutions, though in general only Lipshitz, are smooth in a neighbourhood of the global minimizer. Our result generalizes the result of E, Khanin, Mazel and Sinai (\cite{EKMS00}) to dimension $d\ge 2$, and extends the result of Iturriaga and Khanin in \cite{IK03}.
  • We report the observation and physical characterization of the possible dwarf planet \UZ\ ("DeeDee"), a dynamically detached trans-Neptunian object discovered at 92 AU. This object is currently the second-most distant known trans-Neptunian object with reported orbital elements, surpassed in distance only by the dwarf planet Eris. The object was discovered with an $r$-band magnitude of 23.0 in data collected by the Dark Energy Survey between 2014 and 2016. Its 1140-year orbit has $(a,e,i) = (109~\mathrm{AU}, 0.65, 26.8^{\circ})$. It will reach its perihelion distance of 38 AU in the year 2142. Integrations of its orbit show it to be dynamically stable on Gyr timescales, with only weak interactions with Neptune. We have performed followup observations with ALMA, using 3 hours of on-source integration time to measure the object's thermal emission in the Rayleigh-Jeans tail. The signal is detected at 7$\sigma$ significance, from which we determine a $V$-band albedo of $13.1^{+3.3}_{-2.4}\mathrm{(stat)}^{+2.0}_{-1.4}\mathrm{(sys)}$ percent and a diameter of $635^{+57}_{-61}\mathrm{(stat)}^{+32}_{-39}\mathrm{(sys)}$~km, assuming a spherical body with uniform surface properties.
  • A residual-networks family with hundreds or even thousands of layers dominates major image recognition tasks, but building a network by simply stacking residual blocks inevitably limits its optimization ability. This paper proposes a novel residual-network architecture, Residual networks of Residual networks (RoR), to dig the optimization ability of residual networks. RoR substitutes optimizing residual mapping of residual mapping for optimizing original residual mapping. In particular, RoR adds level-wise shortcut connections upon original residual networks to promote the learning capability of residual networks. More importantly, RoR can be applied to various kinds of residual networks (ResNets, Pre-ResNets and WRN) and significantly boost their performance. Our experiments demonstrate the effectiveness and versatility of RoR, where it achieves the best performance in all residual-network-like structures. Our RoR-3-WRN58-4+SD models achieve new state-of-the-art results on CIFAR-10, CIFAR-100 and SVHN, with test errors 3.77%, 19.73% and 1.59%, respectively. RoR-3 models also achieve state-of-the-art results compared to ResNets on ImageNet data set.
  • We characterize when a finite lattice is distributive by the existences of some particular classes of Koszul filtrations.
  • We improve the global Nekhoroshev stability for analytic quasi-convex nearly integrable Hamiltonian systems. The new stability result is optimal, as it matches the fastest speed of Arnold diffusion.
  • We report observations of resolved C2H emission rings within the gas-rich protoplanetary disks of TW Hya and DM Tau using the Atacama Large Millimeter Array (ALMA). In each case the emission ring is found to arise at the edge of the observable disk of mm-sized grains (pebbles) traced by (sub)mm-wave continuum emission. In addition, we detect a C3H2 emission ring with an identical spatial distribution to C2H in the TW Hya disk. This suggests that these are hydrocarbon rings (i.e. not limited to C2H). Using a detailed thermo-chemical model we show that reproducing the emission from C2H requires a strong UV field and C/O > 1 in the upper disk atmosphere and outer disk, beyond the edge of the pebble disk. This naturally arises in a disk where the ice-coated dust mass is spatially stratified due to the combined effects of coagulation, gravitational settling and drift. This stratification causes the disk surface and outer disk to have a greater permeability to UV photons. Furthermore the concentration of ices that transport key volatile carriers of oxygen and carbon in the midplane, along with photochemical erosion of CO, leads to an elemental C/O ratio that exceeds unity in the UV-dominated disk. Thus the motions of the grains, and not the gas, lead to a rich hydrocarbon chemistry in disk surface layers and in the outer disk midplane.
  • We propose a novel supervised learning technique for summarizing videos by automatically selecting keyframes or key subshots. Casting the problem as a structured prediction problem on sequential data, our main idea is to use Long Short-Term Memory (LSTM), a special type of recurrent neural networks to model the variable-range dependencies entailed in the task of video summarization. Our learning models attain the state-of-the-art results on two benchmark video datasets. Detailed analysis justifies the design of the models. In particular, we show that it is crucial to take into consideration the sequential structures in videos and model them. Besides advances in modeling techniques, we introduce techniques to address the need of a large number of annotated data for training complex learning models. There, our main idea is to exploit the existence of auxiliary annotated video datasets, albeit heterogeneous in visual styles and contents. Specifically, we show domain adaptation techniques can improve summarization by reducing the discrepancies in statistical properties across those datasets.
  • Video summarization has unprecedented importance to help us digest, browse, and search today's ever-growing video collections. We propose a novel subset selection technique that leverages supervision in the form of human-created summaries to perform automatic keyframe-based video summarization. The main idea is to nonparametrically transfer summary structures from annotated videos to unseen test videos. We show how to extend our method to exploit semantic side information about the video's category/genre to guide the transfer process by those training videos semantically consistent with the test input. We also show how to generalize our method to subshot-based summarization, which not only reduces computational costs but also provides more flexible ways of defining visual similarity across subshots spanning several frames. We conduct extensive evaluation on several benchmarks and demonstrate promising results, outperforming existing methods in several settings.
  • We report a test of the universality of free fall (UFF) by comparing the gravity acceleration of the $^{87}$Rb atoms in $m_F=+1$ versus that in $m_F=-1$, where the corresponding spin orientations are opposite. A Mach-Zehnder-type atom interferometer is exploited to sequentially measure the free fall acceleration of the atoms in these two magnetic sublevels, and the resultant E$\rm{\ddot{o}}$tv$\rm{\ddot{o}}$s ratio is ${\eta _S} =(0.2\pm1.2)\times 10^{-7}$. This also gives an upper limit of $1.1\times 10^{-21}$ GeV/m for possible gradient field of the spacetime torsion. The interferometer using atoms in $m_F=\pm 1$ is highly sensitive to the magnetic field inhomogeneity, and a double differential measurement method is developed to alleviate the inhomogeneity influence. Moreover, a proof experiment by modulating the magnetic field is performed, which validates the alleviation of the inhomogeneity influence in our test.
  • An unsolved problem in step-wise core-accretion planet formation is that rapid radial drift in gas-rich protoplanetary disks should drive mm-/meter-sized particles inward to the central star before large bodies can form. One promising solution is to confine solids within small scale structures. Here we investigate dust structures in the (sub)mm continuum emission of four disks (TW Hya, HL Tau, HD 163296 and DM Tau), a sample of disks with the highest spatial resolution ALMA observations to date. We retrieve the surface brightness distributions using synthesized images and fitting visibilities with analytical functions. We find that the continuum emission of the four disks is ~axi-symmetric but rich in 10-30AU-sized radial structures, possibly due to physical gaps, surface density enhancements or localized dust opacity variations within the disks. These results suggest that small scale axi-symmetric dust structures are likely to be common, as a result of ubiquitous processes in disk evolution and planet formation. Compared with recent spatially resolved observations of CO snowlines in these same disks, all four systems show enhanced continuum emission from regions just beyond the CO condensation fronts, potentially suggesting a causal relationship between dust growth/trapping and snowlines.