arXiv subjects

Jie Song

Publications and source records attributed to Jie Song.

At least 253 records · Page 14Linked to original sources

arXiv2019

Customizing Student Networks From Heterogeneous Teachers via Adaptive Knowledge Amalgamation

A massive number of well-trained deep networks have been released by developers online. These networks may focus on different tasks and in many cases are optimized for different datasets. In this paper, we study how to exploit such heterogeneous pre-trained networks, known as teachers, so as to train a customized student network that tackles a set of selective tasks defined by the user. We assume no human annotations are available, and each teacher may be either single- or multi-task. To this end, we introduce a dual-step strategy that first extracts the task-specific knowledge from the heterogeneous teachers sharing the same sub-task, and then amalgamates the extracted knowledge to build the student network. To facilitate the training, we employ a selective learning scheme where, for each unlabelled sample, the student learns adaptively from only the teacher with the least prediction ambiguity. We evaluate the proposed approach on several datasets and experimental results demonstrate that the student, learned by such adaptive knowledge amalgamation, achieves performances even better than those of the teachers.

cs.CV↗

arXiv2019

Investigation of wind pressures on tall building under interference effects using machine learning techniques

Interference effects of tall buildings have attracted numerous studies due to the boom of clusters of tall buildings in megacities. To fully understand the interference effects of buildings, it often requires a substantial amount of wind tunnel tests. Limited wind tunnel tests that only cover part of interference scenarios are unable to fully reveal the interference effects. This study used machine learning techniques to resolve the conflicting requirement between limited wind tunnel tests that produce unreliable results and a completed investigation of the interference effects that is costly and time-consuming. Four machine learning models including decision tree, random forest, XGBoost, generative adversarial networks (GANs), were trained based on 30% of a dataset to predict both mean and fluctuating pressure coefficients on the principal building. The GANs model exhibited the best performance in predicting these pressure coefficients. A number of GANs models were then trained based on different portions of the dataset ranging from 10% to 90%. It was found that the GANs model based on 30% of the dataset is capable of predicting both mean and fluctuating pressure coefficients under unseen interference conditions accurately. By using this GANs model, 70% of the wind tunnel test cases can be saved, largely alleviating the cost of this kind of wind tunnel testing study.

cs.LG↗

arXiv2019

Squeezing enhanced atom-cavity interaction in coupled cavities with high dissipation rates

The realization of the strong coupling regime is requisite for implementing quantum information tasks. Here, a method for enhancing the atom-field coupling in highly dissipative coupled cavities is proposed. By introducing parametric squeezing into the primary cavity which is only virtually excited under specific parametric conditions, coupling enhancement between atom and auxiliary cavity is realized for proper squeezing parameters. This enables the system to be robust against large cavity decay and atomic spontaneous emission. The observation of vacuum Rabi oscillations show that the originally weak-coupled system can be enhanced into effective strong coupling regime.

quant-ph↗

arXiv2019

Unpaired Pose Guided Human Image Generation

This paper studies the task of full generative modelling of realistic images of humans, guided only by coarse sketch of the pose, while providing control over the specific instance or type of outfit worn by the user. This is a difficult problem because input and output domain are very different and direct image-to-image translation becomes infeasible. We propose an end-to-end trainable network under the generative adversarial framework, that provides detailed control over the final appearance while not requiring paired training data and hence allows us to forgo the challenging problem of fitting 3D poses to 2D images. The model allows to generate novel samples conditioned on either an image taken from the target domain or a class label indicating the style of clothing (e.g., t-shirt). We thoroughly evaluate the architecture and the contributions of the individual components experimentally. Finally, we show in a large scale perceptual study that our approach can generate realistic looking images and that participants struggle in detecting fake images versus real samples, especially if faces are blurred.

cs.CV↗

arXiv2019

Resonant-interaction-induced Rydberg antiblockade

Different from the conventional Rydberg antiblockade (RAB) regime that either requires weak Rydberg-Rydberg interaction (RRI), or compensates Rydberg-Rydberg interaction (RRI)-induced energy shift by introducing dispersive interactions, we show that RAB regime can be achieved by resonantly driving the transitions between ground state and Rydberg state under strong RRI. The Rabi frequencies are of small amplitude and time-dependent harmonic oscillation, which plays a critical role for the presented RAB. The proposed unconventional RAB regime is used to construct high-fidelity controlled-Z (CZ) gate and controlled-not (CNOT) gate in one step. Each atom requires single external driving. And the atomic addressability is not required for the presented unconventional RAB, which would simplify experimental complexity and reduce resource consumption.

quant-ph↗

arXiv2019

Enhancement of coherent dipole coupling between two atoms via squeezing a cavity mode

We propose a theoretical method to enhance the coherent dipole coupling between two atoms in an optical cavity via parametrically squeezing the cavity mode. In the present scheme, conditions for coherent coupling are derived in detail and diverse dynamics of the system can be obtained by regulating system parameters. In the presence of environmental noise, an auxiliary squeezed field is employed to suppress, and even completely eliminate the additional noise induced by squeezing. In addition, we demonstrate that our scheme enables the effective suppression of atomic spontaneous emission. The results in our investigation could be used for diverse applications in quantum technologies.

quant-ph↗

arXiv2019

Probery: A Probability-based Incomplete Query Optimization for Big Data

Nowadays, query optimization has been highly concerned in big data management, especially in NoSQL databases. Approximate queries boost query performance by loss of accuracy, for example, sampling approaches trade off query completeness for efficiency. Different from them, we propose an uncertainty of query completeness, called Probability of query Completeness (PC for short). PC refers to the possibility that query results contain all satisfied records. For example PC=0.95, it guarantees that there are no more than 5 incomplete queries among 100 ones, but not guarantees how incomplete they are. We trade off PC for query performance, and experiments show that a small loss of PC doubles query performance. The proposed Probery (PROBability-based data quERY) adopts the uncertainty of query completeness to accelerate OLTP queries. This paper illustrates the data and probability models, the probability based data placement and query processing, and the Apache Drill-based implementation of Probery. In experiments, we first prove that the percentage of complete queries is larger than the given PC confidence for various cases, namely that the PC guarantee is validate. Then Probery is compared with Drill, Impala and Hive in terms of query performance. The results indicate that Drill-based Probery performs as fast as Drill with complete query, while averagely 1.8x, 1.3x and 1.6x faster than Drill, Impala and Hive with possible complete query, respectively.

cs.DB↗

arXiv2018

End-to-end Learning for Graph Decomposition

We propose a novel end-to-end trainable framework for the graph decomposition problem. The minimum cost multicut problem is first converted to an unconstrained binary cubic formulation where cycle consistency constraints are incorporated into the objective function. The new optimization problem can be viewed as a Conditional Random Field (CRF) in which the random variables are associated with the binary edge labels of the initial graph and the hard constraints are introduced in the CRF as high-order potentials. The parameters of a standard Neural Network and the fully differentiable CRF are optimized in an end-to-end manner. Furthermore, our method utilizes the cycle constraints as meta-supervisory signals during the learning of the deep feature representations by taking the dependencies between the output random variables into account. We present analyses of the end-to-end learned representations, showing the impact of the joint training, on the task of clustering images of MNIST. We also validate the effectiveness of our approach both for the feature learning and the final clustering on the challenging task of real-world multi-person pose estimation.

cs.CV↗

arXiv2018

Shortcut Scheme for One-Step Implementation of a Three-Qubit Nonadiabatic Holonomic Gate

Nonadiabatic holonomic quantum computation has attracted increasing interest because of its robustness. In this paper, based on the shortcuts to adiabaticity (STA), we propose a scheme to construct a three-qubit nonadiabatic holonomic gate in a cavity quantum electrodynamics system. In the scheme, the three-qubit holonomic gate is implemented in a single step. Besides, numerical simulations show that the constructed holonomic quantum gate holds the characteristics of high-fidelity, decoherence-resistance and error-immunity.

quant-ph↗

arXiv2018

Amalgamating Knowledge towards Comprehensive Classification

With the rapid development of deep learning, there have been an unprecedentedly large number of trained deep network models available online. Reusing such trained models can significantly reduce the cost of training the new models from scratch, if not infeasible at all as the annotations used for the training original networks are often unavailable to public. We propose in this paper to study a new model-reusing task, which we term as \emph{knowledge amalgamation}. Given multiple trained teacher networks, each of which specializes in a different classification problem, the goal of knowledge amalgamation is to learn a lightweight student model capable of handling the comprehensive classification. We assume no other annotations except the outputs from the teacher models are available, and thus focus on extracting and amalgamating knowledge from the multiple teachers. To this end, we propose a pilot two-step strategy to tackle the knowledge amalgamation task, by learning first the compact feature representations from teachers and then the network parameters in a layer-wise manner so as to build the student model. We apply this approach to four public datasets and obtain very encouraging results: even without any human annotation, the obtained student model is competent to handle the comprehensive classification task and in most cases outperforms the teachers in individual sub-tasks.

cs.CV↗

arXiv2018

Elimination of basal-plane stacking faults in semipolar/nonpolar GaN and light emitting diodes heteroepitaxially grown on sapphire substrates

High quality semipolar and nonpolar GaN is crucial in achieving high-performance GaN-based optoelectronic devices, yet it has been very challenging to achieve large-area wafers that are free of basal-plane stacking faults (BSFs). In this work, we report an approach to prepare large-area, stacking-fault-free (SF-free) semipolar GaN on (4-inch) sapphire substrates. A root cause of the formation of BSFs is the emergence of N-polar (000-1) facets during semipolar and non-polar heteroepitaxy. Invoking the concept of kinetic Wulff plot, we succeeded in suppressing the occurrence of N-polar GaN (000-1) facets, and consequently in eliminating the stacking faults generated in (000-1) basal-planes. The result was confirmed by transmission electron microscopy, cathodoluminescence, and low-temperature photoluminescence characterizations. Furthermore, InGaN light emitting diodes with promising characteristics have been produced on the SF-free semipolar (20-21) GaN on sapphire substrates. Our work opens up a new insight about the heteroepitaxial growth of nonpolar/semipolar GaN and provides an approach of producing SF-free nonpolar/semipolar GaN material over large-area wafers which will create new opportunities in GaN optoelectronic and microelectronic research.

cond-mat.mtrl-sci↗

arXiv2018

Precise quantum control via unsharp measurements and feedback operations

In this paper, we propose a scheme to eliminate the influence of noises on system dynamics, by means of a sequential unsharp measurements and unitary feedback operations. The unsharp measurements are carried out periodically during system evolution, while the feedback operations are well designed based on the eigenstates of the density matrices of the exact (noiseless) dynamical states and its corresponding post-measurement states. For illustrative examples, we show that the dynamical trajectory errors caused by both static and non-static noises are successfully eliminated in typical two-level and multi-level systems, i.e., the high-fidelity quantum dynamics can be maintained. Furthermore, we discuss the influence of noise strength and measurement strength on the degree of precise quantum control. Crucially, the measurement-feedback scheme is quite universal in that it can be applied to precise quantum control for any dimension systems. Thus, it naturally finds extensive applications in quantum information processing.

quant-ph↗

arXiv2018

Selective Zero-Shot Classification with Augmented Attributes

In this paper, we introduce a selective zero-shot classification problem: how can the classifier avoid making dubious predictions? Existing attribute-based zero-shot classification methods are shown to work poorly in the selective classification scenario. We argue the under-complete human defined attribute vocabulary accounts for the poor performance. We propose a selective zero-shot classifier based on both the human defined and the automatically discovered residual attributes. The proposed classifier is constructed by firstly learning the defined and the residual attributes jointly. Then the predictions are conducted within the subspace of the defined attributes. Finally, the prediction confidence is measured by both the defined and the residual attributes. Experiments conducted on several benchmarks demonstrate that our classifier produces a superior performance to other methods under the risk-coverage trade-off metric.

cs.CV↗

arXiv2018

Transductive Unbiased Embedding for Zero-Shot Learning

Most existing Zero-Shot Learning (ZSL) methods have the strong bias problem, in which instances of unseen (target) classes tend to be categorized as one of the seen (source) classes. So they yield poor performance after being deployed in the generalized ZSL settings. In this paper, we propose a straightforward yet effective method named Quasi-Fully Supervised Learning (QFSL) to alleviate the bias problem. Our method follows the way of transductive learning, which assumes that both the labeled source images and unlabeled target images are available for training. In the semantic embedding space, the labeled source images are mapped to several fixed points specified by the source categories, and the unlabeled target images are forced to be mapped to other points specified by the target categories. Experiments conducted on AwA2, CUB and SUN datasets demonstrate that our method outperforms existing state-of-the-art approaches by a huge margin of 9.3~24.5% following generalized ZSL settings, and by a large margin of 0.2~16.2% following conventional ZSL settings.

cs.CV↗

arXiv2018

Cross-modal Deep Variational Hand Pose Estimation

The human hand moves in complex and high-dimensional ways, making estimation of 3D hand pose configurations from images alone a challenging task. In this work we propose a method to learn a statistical hand model represented by a cross-modal trained latent space via a generative deep neural network. We derive an objective function from the variational lower bound of the VAE framework and jointly optimize the resulting cross-modal KL-divergence and the posterior reconstruction objective, naturally admitting a training regime that leads to a coherent latent space across multiple modalities such as RGB images, 2D keypoint detections or 3D hand configurations. Additionally, it grants a straightforward way of using semi-supervision. This latent space can be directly used to estimate 3D hand poses from RGB images, outperforming the state-of-the art in different settings. Furthermore, we show that our proposed method can be used without changes on depth images and performs comparably to specialized methods. Finally, the model is fully generative and can synthesize consistent pairs of hand configurations across modalities. We evaluate our method on both RGB and depth datasets and analyze the latent space qualitatively.

cs.CV↗

arXiv2018

Accelerated and noise-resistant generation of a high-fidelity steady-state entanglement with Rydberg atoms

Based on Lyapunov control, a scheme is proposed to accelerate the dissipation dynamics for the generation of high-fidelity entanglement between two Rydberg atoms in the context of cavity quantum electrodynamics (QED). We first use the quantum Zeno dynamics and Rydberg antiblockade to find a unique steady state (two-atom singlet state) for the system. Then, applying additional coherent control (ACC) fields to improve the evolution speed of the dissipative system. The ACC fields are designed based on the target state and they vanish gradually along with increasing of the fidelity thus the system is guaranteed to be finally stable. Besides, the current accelerated scheme is checked to be robust against systematic and amplitude-noise errors.

quant-ph↗

arXiv2018

Invariant-based inverse engineering for fluctuation transfer between membranes in an optomechanical cavity system

In this paper, by invariant-based inverse engineering, we design classical driving fields to transfer quantum fluctuations between two suspended membranes in an optomechanical cavity system. The transfer can be quickly attained through a non-adiabatic evolution path determined by a so-called dynamical invariant. Such an evolution path allows one to optimize the occupancies of the unstable "intermediate" states thus the influence of cavity decays can be suppressed. Numerical simulation demonstrates that a perfect fluctuation transfer between two membranes can be rapidly achieved in one step, and the transfer is robust to both the amplitude noises and cavity decays.

quant-ph↗

arXiv2018

Driving many distant atoms into high-fidelity steady state entanglement via Lyapunov control

Based on Lyapunov control theory in closed and open systems, we propose a scheme to generate W state of many distant atoms in the cavity-fiber-cavity system. In the closed system, the W state is generated successfully even when the coupling strength between the cavity and fiber is extremely weak. In the presence of atomic spontaneous emission or cavity and fiber decay, the photon-measurement and quantum feedback approaches are proposed to improve the fidelity, which enable efficient generation of high-fidelity W state in the case of large dissipation. Furthermore, the time-optimal Lyapunov control is investigated to shorten the evolution time and improve the fidelity in open systems.

quant-ph↗