arXiv subjects

Yan Zhu

Publications and source records attributed to Yan Zhu.

At least 127 records · Page 7Linked to original sources

arXiv2016

Semantic Amodal Segmentation

Common visual recognition tasks such as classification, object detection, and semantic segmentation are rapidly reaching maturity, and given the recent rate of progress, it is not unreasonable to conjecture that techniques for many of these problems will approach human levels of performance in the next few years. In this paper we look to the future: what is the next frontier in visual recognition? We offer one possible answer to this question. We propose a detailed image annotation that captures information beyond the visible pixels and requires complex reasoning about full scene structure. Specifically, we create an amodal segmentation of each image: the full extent of each region is marked, not just the visible pixels. Annotators outline and name all salient regions in the image and specify a partial depth order. The result is a rich scene structure, including visible and occluded portions of each region, figure-ground edge information, semantic labels, and object overlap. We create two datasets for semantic amodal segmentation. First, we label 500 images in the BSDS dataset with multiple annotators per image, allowing us to study the statistics of human annotations. We show that the proposed full scene annotation is surprisingly consistent between annotators, including for regions and edges. Second, we annotate 5000 images from COCO. This larger dataset allows us to explore a number of algorithmic ideas for amodal segmentation and depth ordering. We introduce novel metrics for these tasks, and along with our strong baselines, define concrete new challenges for the community.

cs.CV↗

arXiv2016

Better Computer Go Player with Neural Network and Long-term Prediction

Competing with top human players in the ancient game of Go has been a long-term goal of artificial intelligence. Go's high branching factor makes traditional search techniques ineffective, even on leading-edge hardware, and Go's evaluation function could change drastically with one stone change. Recent works [Maddison et al. (2015); Clark & Storkey (2015)] show that search is not strictly necessary for machine Go players. A pure pattern-matching approach, based on a Deep Convolutional Neural Network (DCNN) that predicts the next move, can perform as well as Monte Carlo Tree Search (MCTS)-based open source Go engines such as Pachi [Baudis & Gailly (2012)] if its search budget is limited. We extend this idea in our bot named darkforest, which relies on a DCNN designed for long-term predictions. Darkforest substantially improves the win rate for pattern-matching approaches against MCTS-based approaches, even with looser search budgets. Against human players, the newest versions, darkfores2, achieve a stable 3d level on KGS Go Server as a ranked bot, a substantial improvement upon the estimated 4k-5k ranks for DCNN reported in Clark & Storkey (2015) based on games against other machine players. Adding MCTS to darkfores2 creates a much stronger player named darkfmcts3: with 5000 rollouts, it beats Pachi with 10k rollouts in all 250 games; with 75k rollouts it achieves a stable 5d level in KGS server, on par with state-of-the-art Go AIs (e.g., Zen, DolBaram, CrazyStone) except for AlphaGo [Silver et al. (2016)]; with 110k rollouts, it won the 3rd place in January KGS Go Tournament.

cs.LG↗

arXiv2016

On Layered Erasure Interference Channels without CSI at Transmitters

This paper studies a layered erasure model for two-user interference channels, which can be viewed as a simplified version of Gaussian fading interference channel. It is assumed that channel state information~(CSI) is only available at receivers but not at transmitters. Under such assumption, an outer bound is derived for the capacity region of such interference channel. The new outer bound is tight in many circumstances. For the remaining open cases, the outer bound extends previous results.

cs.IT↗

arXiv2015

Weak and strong coupling equilibration in nonabelian gauge theories

We present a direct comparison studying equilibration through kinetic theory at weak coupling and through holography at strong coupling in the same set-up. The set-up starts with a homogeneous thermal state, which then smoothly transitions through an out-of-equilibrium phase to an expanding system undergoing boost-invariant flow. This first apples-to-apples comparison of equilibration provides a benchmark for similar equilibration processes in heavy-ion collisions, where the equilibration mechanism is still under debate. We find that results at weak and strong coupling can be smoothly connected by simple, empirical power-laws for the viscosity, equilibration time and entropy production of the system.

hep-th↗

arXiv2015

Relative t-designs in binary Hamming association scheme H(n,2)

A relative t-design in the binary Hamming association schemes H(n,2) is equivalent to a weighted regular t-wise balanced design, i.e., certain combinatorial t-design which allow different sizes of blocks and a weight function on blocks. In this paper, we study relative t-designs in H(n,2), putting emphasis on Fisher type inequalities and the existence of tight relative t-designs. We mostly consider relative t-designs on two shells. We prove that if the weight function is constant on each shell of a relative t-design on two shells then the subset in each shell must be a combinatorial (t-1)-design. This is a generalization of the result of Kageyama who proved this under the stronger assumption that the weight function is constant on the whole block set. Using this, we define tight relative t-designs for odd t, and a strong restriction on the possible parameters of tight relative t-designs in H(n,2). We obtained a new family of such tight relative t-designs, which were unnoticed before. We will give a list of feasible parameters of such relative 3-designs with n up to 100, and then we discuss the existence and/or the non-existence of such tight relative 3-designs. We also discuss feasible parameters of tight relative 4-designs on two shells in H(n,2) with n up 50. In this study we come up with the connection on the topics of classical design theory, such as symmetric 2-designs (in particular 2-(4u-1,2u-1,u-1) Hadamard designs) and Driessen's result on the non-existence of certain 3-designs. We believe the Problem 1 and Problem 2 presented in Section 5.2 open a new way to study relative t-designs in H(n,2). We conclude our paper listing several open problems.

math.CO↗

arXiv2015

Isotropization and hydrodynamization in weakly coupled heavy-ion collisions

We numerically solve 2+1D effective kinetic theory of weak coupling QCD under longitudinal expansion relevant for early stages of heavy-ion collisions. We find agreement with viscous hydrodynamics and classical Yang-Mills simulations in the regimes where they are applicable. By choosing initial conditions that are motivated by color-glass-condensate framework we find that for Q=2GeV and $α_s$=0.3 the system is approximately described by viscous hydrodynamics well before $τ\lesssim 1.0$ fm/c.

hep-ph↗

arXiv2015

More on spherical designs of harmonic index $t$

A finite subset $Y$ on the unit sphere $S^{n-1} \subseteq \mathbb{R}^n$ is called a spherical design of harmonic index $t$, if the following condition is satisfied: $\sum_{\mathbf{x}\in Y}f(\mathbf{x})=0$ for all real homogeneous harmonic polynomials $f(x_1,\ldots,x_n)$ of degree $t$. Also, for a subset $T$ of $\mathbb{N} = \{1,2,\cdots \}$, a finite subset $Y\subset S^{n-1}$ is called a spherical design of harmonic index $T,$ if $\sum_{\mathbf{x}\in Y}f(\mathbf{x})=0$ is satisfied for all real homogeneous harmonic polynomials $f(x_1,\ldots,x_n)$ of degree $k$ with $k\in T$. In the present paper we first study Fisher type lower bounds for the sizes of spherical designs of harmonic index $t$ (or for harmonic index $T$). We also study 'tight' spherical designs of harmonic index $t$ or index $T$. Here 'tight' means that the size of $Y$ attains the lower bound for this Fisher type inequality. The classification problem of tight spherical designs of harmonic index $t$ was started by Bannai-Okuda-Tagami (2015), and the case $t = 4$ was completed by Okuda-Yu (2015+). In this paper we show the classification (non-existence) of tight spherical designs of harmonic index 6 and 8, as well as the asymptotic non-existence of tight spherical designs of harmonic index $2e$ for general $e\geq 3$. We also study the existence problem for tight spherical designs of harmonic index $T$ for some $T$, in particular, including index $T = \{8,4\}$. We use (i) the linear programming method by Delsarte, (ii) the detailed information on the locations of the zeros as well as the local minimum values of Gegenbauer polynomials, (iii) the generalization by Hiroshi Nozaki of the Larman-Rogers-Seidel theorem on $2$-distance sets to $s$-distance sets, (iv) the theory of elliptic diophantine equations, and (v) the semidefinite programming method of eliminating some $2$-angular line systems for small dimensions.

math.CO↗

arXiv2015

Jet propagation within a Linearized Boltzmann Transport Model

A Linear Boltzmann Transport (LBT) model has been developed for the study of jet propagation inside a quark-gluon plasma. Both leading and thermal recoiled partons are transported according to the Boltzmann equations to account for jet-induced medium excitations. In this talk, we present our study within the LBT model in which we implement the complete set of elastic parton scattering processes. We investigate elastic parton energy loss and their energy and length dependence. We further investigate elastic energy loss and transverse shape of reconstructed jets. Contributions from the recoiled thermal partons are found to have significant influences on the jet energy loss and transverse profile.

nucl-th↗

arXiv2015

On the infrared behavior of the shear spectral function in hot Yang-Mills theory

We revisit the determination of the two-loop spectral function in the shear channel of hot Yang-Mills theory. Correcting a technical error in an earlier computation and extending the result with a leading order Hard Thermal Loop resummation is seen to improve the infrared behavior of the quantity significantly. This makes it possible to straightforwardly use the result in the corresponding imaginary time correlator and the shear sum rule.

hep-ph↗

arXiv2014

On parallel multisplitting methods for non-Hermitian positive definite linear systems

To solve non-Hermitian linear system Ax=b on parallel and vector machines, some paralell multisplitting methods are considered. In this work, in particular: i) We establish the convergence results of the paralell multisplitting methods, together with its relaxed version, some of which can be regarded as generalizations of analogous results for the Hermitian positive definite case; ii) We extend the positive-definite and skew-Hermitian splitting (PSS) method methods in [{\em SIAM J. Sci. Comput.}, 26:844--863, 2005] to the parallel PSS methods and propose the corresponding convergence results.

math.NA↗

arXiv2014

MITEoR: A Scalable Interferometer for Precision 21 cm Cosmology

We report on the MIT Epoch of Reionization (MITEoR) experiment, a pathfinder low-frequency radio interferometer whose goal is to test technologies that improve the calibration precision and reduce the cost of the high-sensitivity 3D mapping required for 21 cm cosmology. MITEoR accomplishes this by using massive baseline redundancy, which enables both automated precision calibration and correlator cost reduction. We demonstrate and quantify the power and robustness of redundancy for scalability and precision. We find that the calibration parameters precisely describe the effect of the instrument upon our measurements, allowing us to form a model that is consistent with $χ^2$ per degree of freedom < 1.2 for as much as 80% of the observations. We use these results to develop an optimal estimator of calibration parameters using Wiener filtering, and explore the question of how often and how finely in frequency visibilities must be reliably measured to solve for calibration coefficients. The success of MITEoR with its 64 dual-polarization elements bodes well for the more ambitious Hydrogen Epoch of Reionization Array (HERA) project and other next-generation instruments, which would incorporate many identical or similar technologies.

astro-ph.IM↗

arXiv2014

Jet quenching and $γ$-jet correlation in high-energy heavy-ion collisions

Medium modification of $γ$-tagged jets in high-energy heavy-ion collisions is investigated within a linearized Boltzmann transport model which includes both elastic parton scattering and induced gluon emission. In Pb+Pb collisions at $\sqrt{s}=2.76$ TeV, a $γ$-tagged jet is seen to lose 15\% of its energy at 0-10\% central collisions. Simulations also point to a sizable azimuthal angle broadening of $γ$-tagged jets at the tail of a distribution which should be measurable when experimental errors are significantly reduced. An enhancement at large $z_\text{jet}=p_L/E_{\text{jet}}$ in jet fragmentation function at the Large Hadron Collider (LHC) can be attributed to the dominance of leading particles in the reconstructed jet. A $γ-$tagged jet fragmentation function is shown to be more sensitive to jet quenching, therefore a better probe of the jet transport parameter.

hep-ph↗

arXiv2013

Mapping our Universe in 3D with MITEoR

Mapping our universe in 3D by imaging the redshifted 21 cm line from neutral hydrogen has the potential to overtake the cosmic microwave background as our most powerful cosmological probe, because it can map a much larger volume of our Universe, shedding new light on the epoch of reionization, inflation, dark matter, dark energy, and neutrino masses. We report on MITEoR, a pathfinder low-frequency radio interferometer whose goal is to test technologies that greatly reduce the cost of such 3D mapping for a given sensitivity. MITEoR accomplishes this by using massive baseline redundancy both to enable automated precision calibration and to cut the correlator cost scaling from N^2 to NlogN, where N is the number of antennas. The success of MITEoR with its 64 dual-polarization elements bodes well for the more ambitious HERA project, which would incorporate many identical or similar technologies using an order of magnitude more antennas, each with dramatically larger collecting area.

astro-ph.IM↗

arXiv2013

Medium Modification of γ-jets in High-energy Heavy-ion Collisions

Medium modification of γ-tagged jets in high-energy heavy-ion collisions is investigated within a Linearized Boltzmann Transport model for jet propagation that includes both elastic parton scattering and induced gluon emission. Inclusion of recoiled medium partons in the reconstruction of partonic jets is found to significantly reduce the net jet energy loss. Experimental data on γ-jet asymmetry and survival rate in Pb + Pb collisions at \sqrt{s}=2.76 TeV can be reproduced. Medium modifications of reconstructed jet fragmentation function, transverse profile and energy flow outside the jet-cone are found to be sizable especially for γ-tagged jets with small values of x=p_T^{jet}/p_T^γ.

hep-ph↗

arXiv2013

On the Perturbative Evaluation of Thermal Green's Functions in the Bulk and Shear Channels of Yang-Mills Theory

In this PhD thesis, I will review recent progress in perturbative studies of energy momentum tensor correlators in high-temperature Yang-Mills theory. After briefly introducing the necessary tools and physical motivation, I proceed to discuss the machinery developed for the extraction of next-to-leading order Operator Product Expansions and thermal spectral functions and to introduce the results obtained in the bulk and shear channels of Yang-Mills theory. Particular emphasis is placed on the comparison of the results with recent lattice and gauge/gravity calculations, as well as on discussing their use in extracting the corresponding transport coefficients from Euclidean lattice data.

hep-ph↗

arXiv2013

Tap-Wave-Rub: Lightweight Malware Prevention for Smartphones Using Intuitive Human Gestures

In this paper, we introduce a lightweight permission enforcement approach - Tap-Wave-Rub (TWR) - for smartphone malware prevention. TWR is based on simple human gestures that are very quick and intuitive but less likely to be exhibited in users' daily activities. Presence or absence of such gestures, prior to accessing an application, can effectively inform the OS whether the access request is benign or malicious. Specifically, we present the design of two mechanisms: (1) accelerometer based phone tapping detection; and (2) proximity sensor based finger tapping, rubbing or hand waving detection. The first mechanism is geared for NFC applications, which usually require the user to tap her phone with another device. The second mechanism involves very simple gestures, i.e., tapping or rubbing a finger near the top of phone's screen or waving a hand close to the phone, and broadly appeals to many applications (e.g., SMS). In addition, we present the TWR-enhanced Android permission model, the prototypes implementing the underlying gesture recognition mechanisms, and a variety of novel experiments to evaluate these mechanisms. Our results suggest the proposed approach could be very effective for malware detection and prevention, with quite low false positives and false negatives, while imposing little to no additional burden on the users.

cs.CR↗

arXiv2013

Bulk and shear spectral functions in weakly and strongly coupled Yang-Mills theory

In this talk, we discuss a number of recent calculations aimed at determining the spectral functions corresponding to various components of the energy momentum tensor in high-temperature SU(N) Yang-Mills theory. The computations reviewed include applications of both weak coupling and gauge/gravity techniques, and thus enable one to access different limits of the quantities. The motivation for the work is twofold: On one hand, the results are hoped to aid the eventual nonperturbative extraction of the bulk and shear viscosities from lattice data, while on the other hand they also enable an immediate comparison of the lattice, perturbative and holographic predictions for certain Euclidean correlators.

hep-ph↗

arXiv2012

The shear channel spectral function in hot Yang-Mills theory

We determine a next-to-leading order result for the shear channel thermal spectral function in SU(N) Yang-Mills theory, working in the limit of vanishing external three-momentum. The result is subsequently applied to the evaluation of the corresponding imaginary time correlator, and its use in the context of sum rules is discussed. Our hope is that the calculation will eventually find use in the nonperturbative determination of the shear viscosity of the theory.

hep-ph↗