Searcharxiv⌕ Search

arXiv subjects

Yuan Liu

Publications and source records attributed to Yuan Liu.

At least 469 records · Page 26Linked to original sources
arXiv2019

Multi-granularity Generator for Temporal Action Proposal

Temporal action proposal generation is an important task, aiming to localize the video segments containing human actions in an untrimmed video. In this paper, we propose a multi-granularity generator (MGG) to perform the temporal action proposal from different granularity perspectives, relying on the video visual features equipped with the position embedding information. First, we propose to use a bilinear matching model to exploit the rich local information within the video sequence. Afterwards, two components, namely segment proposal producer (SPP) and frame actionness producer (FAP), are combined to perform the task of temporal action proposal at two distinct granularities. SPP considers the whole video in the form of feature pyramid and generates segment proposals from one coarse perspective, while FAP carries out a finer actionness evaluation for each video frame. Our proposed MGG can be trained in an end-to-end fashion. By temporally adjusting the segment proposals with fine-grained frame actionness information, MGG achieves the superior performance over state-of-the-art methods on the public THUMOS-14 and ActivityNet-1.3 datasets. Moreover, we employ existing action classifiers to perform the classification of the proposals generated by MGG, leading to significant improvements compared against the competing methods for the video detection task.

cs.CV↗
arXiv2018

PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation

This paper addresses the challenge of 6DoF pose estimation from a single RGB image under severe occlusion or truncation. Many recent works have shown that a two-stage approach, which first detects keypoints and then solves a Perspective-n-Point (PnP) problem for pose estimation, achieves remarkable performance. However, most of these methods only localize a set of sparse keypoints by regressing their image coordinates or heatmaps, which are sensitive to occlusion and truncation. Instead, we introduce a Pixel-wise Voting Network (PVNet) to regress pixel-wise unit vectors pointing to the keypoints and use these vectors to vote for keypoint locations using RANSAC. This creates a flexible representation for localizing occluded or truncated keypoints. Another important feature of this representation is that it provides uncertainties of keypoint locations that can be further leveraged by the PnP solver. Experiments show that the proposed approach outperforms the state of the art on the LINEMOD, Occlusion LINEMOD and YCB-Video datasets by a large margin, while being efficient for real-time pose estimation. We further create a Truncation LINEMOD dataset to validate the robustness of our approach against truncation. The code will be avaliable at https://zju-3dv.github.io/pvnet/.

cs.CV↗
arXiv2018

Accretion in Strong Field Gravity with eXTP

In this paper we describe the potential of the enhanced X-ray Timing and Polarimetry (eXTP) mission for studies related to accretion flows in the strong field gravity regime around both stellar-mass and supermassive black-holes. eXTP has the unique capability of using advanced 'spectral-timing-polarimetry' techniques to analyze the rapid variations with three orthogonal diagnostics of the flow and its geometry, yielding unprecedented insight into the inner accreting regions, the effects of strong field gravity on the material within them and the powerful outflows which are driven by the accretion process.

astro-ph.HE↗
arXiv2018

Multiuser Computation Offloading and Downloading for Edge Computing with Virtualization

Mobile-edge computing (MEC) is an emerging technology for enhancing the computational capabilities of mobile devices and reducing their energy consumption via offloading complex computation tasks to the nearby servers. Multiuser MEC at servers is widely realized via parallel computing based on virtualization. Due to finite shared I/O resources, interference between virtual machines (VMs), called I/O interference, degrades the computation performance. In this paper, we study the problem of joint radio-and-computation resource allocation (RCRA) in multiuser MEC systems in the presence of I/O interference. Specifically, offloading scheduling algorithms are designed targeting two system performance metrics: sum offloading throughput maximization and sum mobile energy consumption minimization. Their designs are formulated as non-convex mixed-integer programming problems, which account for latency due to offloading, result downloading and parallel computing. A set of low-complexity algorithms are designed based on a decomposition approach and leveraging classic techniques from combinatorial optimization. The resultant algorithms jointly schedule offloading users, control their offloading sizes, and divide time for communication (offloading and downloading) and computation. They are either optimal or can achieve close-to-optimality as shown by simulation. Comprehensive simulation results demonstrate considering of I/O interference can endow on an offloading controller robustness against the performance-degradation factor.

cs.IT↗
arXiv2018

Thermal Infrared Colorization via Conditional Generative Adversarial Network

Transforming a thermal infrared image into a realistic RGB image is a challenging task. In this paper we propose a deep learning method to bridge this gap. We propose learning the transformation mapping using a coarse-to-fine generator that preserves the details. Since the standard mean squared loss cannot penalize the distance between colorized and ground truth images well, we propose a composite loss function that combines content, adversarial, perceptual and total variation losses. The content loss is used to recover global image information while the latter three losses are used to synthesize local realistic textures. Quantitative and qualitative experiments demonstrate that our approach significantly outperforms existing approaches.

cs.CV↗
arXiv2018

Krylov implicit integration factor discontinuous Galerkin methods on sparse grids for high dimensional reaction-diffusion equations

Computational costs of numerically solving multidimensional partial differential equations (PDEs) increase significantly when the spatial dimensions of the PDEs are high, due to large number of spatial grid points. For multidimensional reaction-diffusion equations, stiffness of the system provides additional challenges for achieving efficient numerical simulations. In this paper, we propose a class of Krylov implicit integration factor (IIF) discontinuous Galerkin (DG) methods on sparse grids to solve reaction-diffusion equations on high spatial dimensions. The key ingredient of spatial DG discretization is the multiwavelet bases on nested sparse grids, which can significantly reduce the numbers of degrees of freedom. To deal with the stiffness of the DG spatial operator in discretizing reaction-diffusion equations, we apply the efficient IIF time discretization methods, which are a class of exponential integrators. Krylov subspace approximations are used to evaluate the large size matrix exponentials resulting from IIF schemes for solving PDEs on high spatial dimensions. Stability and error analysis for the semi-discrete scheme are performed. Numerical examples of both scalar equations and systems in two and three spatial dimensions are provided to demonstrate the accuracy and efficiency of the methods. The stiffness of the reaction-diffusion equations is resolved well and large time step size computations are obtained.

math.NA↗
arXiv2018

Energy Efficiency of Distributed Antenna Systems with Wireless Power Transfer

In this paper, we study energy-efficient resource allocation in distributed antenna system (DAS) with wireless power transfer, where time-division multiple access (TDMA) is adopted for downlink multiuser information transmission. In particular, when a user is scheduled to receive information, other users harvest energy at the same time using the same radio-frequency (RF) signal. We consider two types of energy efficiency (EE) metrics: user-centric EE (UC-EE) and network-centric EE (NC-EE). Our goal is to maximize the UC-EE and NC-EE, respectively, by optimizing the transmission time and power subject to the energy harvesting requirements of the users. For both UC-EE and NC-EE maximization problems, we transform the nonconvex problems into equivalently tractable problems by using suitable mathematical tools and then develop iterative algorithms to find the globally optimal solutions. Simulation results demonstrate the superiority of the proposed methods compared with the benchmark schemes.

cs.IT↗
arXiv2018

Ab Initio Finite Temperature Auxiliary Field Quantum Monte Carlo

We present an \textit{ab initio} auxiliary field quantum Monte Carlo method for studying the electronic structure of molecules, solids, and model Hamiltonians at finite temperature. The algorithm marries the \textit{ab initio} phaseless auxiliary field quantum Monte Carlo algorithm known to produce high accuracy ground state energies of molecules and solids with its finite temperature variant, long used by condensed matter physicists for studying model Hamiltonian phase diagrams, to yield a phaseless, \textit{ab initio} finite temperature method. We demonstrate that the method produces internal energies within chemical accuracy of exact diagonalization results across a wide range of temperatures for H$_{2}$O (STO-3G), C$_{2}$ (STO-6G), the one-dimensional hydrogen chain (STO-6G), and the multi-orbital Hubbard model. Our method effectively controls the phase problem through importance sampling, \textit{often even without invoking the phaseless approximation}, down to temperatures at which the systems studied approach their ground states and may therefore be viewed as exact over wide temperature ranges. This technique embodies a versatile tool for studying the finite temperature phase diagrams of a plethora of systems whose properties cannot be captured by a Hubbard U term alone. Our results moreover illustrate that the severity of the phase problem for model Hamiltonians far exceeds that for many molecules at all of the temperatures studied.

cond-mat.str-el↗
arXiv2018

VSE-ens: Visual-Semantic Embeddings with Efficient Negative Sampling

Jointing visual-semantic embeddings (VSE) have become a research hotpot for the task of image annotation, which suffers from the issue of semantic gap, i.e., the gap between images' visual features (low-level) and labels' semantic features (high-level). This issue will be even more challenging if visual features cannot be retrieved from images, that is, when images are only denoted by numerical IDs as given in some real datasets. The typical way of existing VSE methods is to perform a uniform sampling method for negative examples that violate the ranking order against positive examples, which requires a time-consuming search in the whole label space. In this paper, we propose a fast adaptive negative sampler that can work well in the settings of no figure pixels available. Our sampling strategy is to choose the negative examples that are most likely to meet the requirements of violation according to the latent factors of images. In this way, our approach can linearly scale up to large datasets. The experiments demonstrate that our approach converges 5.02x faster than the state-of-the-art approaches on OpenImages, 2.5x on IAPR-TCI2 and 2.06x on NUS-WIDE datasets, as well as better ranking accuracy across datasets.

cs.CV↗
arXiv2018

An Attention-Based Approach for Single Image Super Resolution

The main challenge of single image super resolution (SISR) is the recovery of high frequency details such as tiny textures. However, most of the state-of-the-art methods lack specific modules to identify high frequency areas, causing the output image to be blurred. We propose an attention-based approach to give a discrimination between texture areas and smooth areas. After the positions of high frequency details are located, high frequency compensation is carried out. This approach can incorporate with previously proposed SISR networks. By providing high frequency enhancement, better performance and visual effect are achieved. We also propose our own SISR network composed of DenseRes blocks. The block provides an effective way to combine the low level features and high level features. Extensive benchmark evaluation shows that our proposed method achieves significant improvement over the state-of-the-art works in SISR.

cs.CV↗
arXiv2018

Charge-then-Forward: Wireless Powered Communication for Multiuser Relay Networks

This paper studies a relay-assisted wireless powered communication network (R-WPCN) consisting of multiple source-destination pairs and a hybrid relay node (HRN). We consider a "charge-then-forward" protocol at the HRN, in which the HRN with constant energy supply first acts as an energy transmitter to charge the sources, and then forwards the information from the sources to their destinations through time division multiple access (TDMA) or frequency division multiple access (FDMA). Processing costs at the wireless-powered sources are taken into account. Our goal is to maximize the sum-rate of all transmission pairs by jointly optimizing the time, frequency and power resources. The formulated optimization problems for both TDMA and FDMA are non-convex. For the TDMA scheme, by appropriate transformation, the problem is reformulated as a convex problem and be optimally solved. For the FDMA case, we find the asymptotically optimal solution in the dual domain. Furthermore, suboptimal algorithms are proposed for both schemes to tradeoff the complexity and performance. Finally, the simulation results validate the effectiveness of the proposed schemes.

cs.IT↗
arXiv2018

Best Vision Technologies Submission to ActivityNet Challenge 2018-Task: Dense-Captioning Events in Videos

This note describes the details of our solution to the dense-captioning events in videos task of ActivityNet Challenge 2018. Specifically, we solve this problem with a two-stage way, i.e., first temporal event proposal and then sentence generation. For temporal event proposal, we directly leverage the three-stage workflow in [13, 16]. For sentence generation, we capitalize on LSTM-based captioning framework with temporal attention mechanism (dubbed as LSTM-T). Moreover, the input visual sequence to the LSTM-based video captioning model is comprised of RGB and optical flow images. At inference, we adopt a late fusion scheme to fuse the two LSTM-based captioning models for sentence generation.

cs.CV↗
arXiv2018

DuReader: a Chinese Machine Reading Comprehension Dataset from Real-world Applications

This paper introduces DuReader, a new large-scale, open-domain Chinese ma- chine reading comprehension (MRC) dataset, designed to address real-world MRC. DuReader has three advantages over previous MRC datasets: (1) data sources: questions and documents are based on Baidu Search and Baidu Zhidao; answers are manually generated. (2) question types: it provides rich annotations for more question types, especially yes-no and opinion questions, that leaves more opportunity for the research community. (3) scale: it contains 200K questions, 420K answers and 1M documents; it is the largest Chinese MRC dataset so far. Experiments show that human performance is well above current state-of-the-art baseline systems, leaving plenty of room for the community to make improvements. To help the community make these improvements, both DuReader and baseline systems have been posted online. We also organize a shared competition to encourage the exploration of more models. Since the release of the task, there are significant improvements over the baselines.

cs.CL↗
arXiv2018

Energy-Efficient SWIPT in IoT Distributed Antenna Systems

The rapid growth of Internet of Things (IoT) dramatically increases power consumption of wireless devices. Simultaneous wireless information and power transfer (SWIPT) is a promising solution for sustainable operation of IoT devices. In this paper, we study energy efficiency (EE) in SWIPT-based distributed antenna system (DAS), where power splitting (PS) is applied at IoT devices to coordinate the energy harvesting (EH) and information decoding (ID) processes by varying transmit power of distributed antenna (DA) ports and PS ratios of IoT devices. In the case of single IoT device, we find the optimal closed-form solution by deriving some useful properties based on Karush-Kuhn-Tucker (KKT) conditions and the solution is no need for numerical iterations. For the case of multiple IoT devices, we propose an efficient suboptimal algorithm to solve the EE maximization problem. Simulation results show that the proposed schemes achieve better EE performance compared with other benchmark schemes in both single and multiple IoT devices cases.

cs.IT↗
arXiv2017

Nintendo Super Smash Bros. Melee: An "Untouchable" Agent

Nintendo's Super Smash Bros. Melee fighting game can be emulated on modern hardware allowing us to inspect internal memory states, such as character positions. We created an AI that avoids being hit by training using these internal memory states and outputting controller button presses. After training on a month's worth of Melee matches, our best agent learned to avoid the toughest AI built into the game for a full minute 74.6% of the time.

cs.AI↗
arXiv2017

Secure Transmission in Linear Multihop Relaying Networks

This paper studies the design and secrecy performance of linear multihop networks, in the presence of randomly distributed eavesdroppers in a large-scale two-dimensional space. Depending on whether there is feedback from the receiver to the transmitter, we study two transmission schemes: on-off transmission (OFT) and non-on-off transmission (NOFT). In the OFT scheme, transmission is suspended if the instantaneous received signal-to-noise ratio (SNR) falls below a given threshold, whereas there is no suspension of transmission in the NOFT scheme. We investigate the optimal design of the linear multiple network in terms of the optimal rate parameters of the wiretap code as well as the optimal number of hops. These design parameters are highly interrelated since more hops reduces the distance of per-hop communication which completely changes the optimal design of the wiretap coding rates. Despite the analytical difficulty, we are able to characterize the optimal designs and the resulting secure transmission throughput in mathematically tractable forms in the high SNR regime. Our numerical results demonstrate that our analytical results obtained in the high SNR regime are accurate at practical SNR values. Hence, these results provide useful guidelines for designing linear multihop networks with targeted physical layer security performance.

cs.IT↗
arXiv2017

Secrecy Rate Maximization with Outage Constraint in Multihop Relaying Networks

In this paper, we study the secure transmission in multihop wireless networks with randomize-and-forward (RaF) relaying, in the presence of randomly distributed eavesdroppers. By considering adaptive encoder with on-off transmission (OFT) scheme, we investigate the optimal design of the wiretap code and routing strategies to maximize the secrecy rate while satisfying the secrecy outage probability (SOP) constraint. We derive the exact expressions for the optimal rate parameters of the wiretap code. Then the secure routing problem is solved by revising the classical Bellman-Ford algorithm. Simulation results are conducted to verify our analysis.

cs.IT↗