Time filter

Source Type

Zhou B.,Shanghai JiaoTong University | Cui Y.,Shanghai JiaoTong University | Tao M.,Shanghai JiaoTong University | Tao M.,Cooperative Medianet Innovation Center
Proceedings of the 1st Workshop on Content Caching and Delivery in Wireless Networks, CCDWN 2016 | Year: 2016

Caching at small base stations (SBSs) has demonstrated significant benefits in alleviating the backhaul requirement in heterogeneous cellular networks (HetNets). While many existing works focus on what contents to cache at each SBS, an equally important but much less investigated problem is what contents to deliver given the cache status and user requests. In this paper, we study the optimal content delivery strategy in cache-enabled HetNets by taking into account the inherent multicast capability of wireless medium. We establish a content-centric request queue model and then formulate a stochastic multicast scheduling problem to jointly minimize the average network delay and power costs. This stochastic optimization problem is an infinite horizon average cost Markov decision process (MDP), which is well known to be challenging. By using relative value iteration algorith- m and the special properties of the request queue dynamics, we characterize some properties of the value function of the MDP. Based on these properties, we show that the optimal multicast scheduling policy, which is adaptive to the request queue state, is of the threshold type. Finally, we propose a low complexity optimal algorithm by exploiting the structural properties of the optimal policy. © 2015 ACM.

Li J.,Beihang University | Tian Y.,Peking University | Tian Y.,Cooperative Medianet Innovation Center | Chen X.,Beihang University | And 2 more authors.
International Journal of Computer Vision | Year: 2016

Detecting conspicuous image content is a challenging task in the field of computer vision. In existing studies, most approaches focus on estimating saliency only with the cues from the input image. However, such “intrinsic” cues are often insufficient to distinguish targets and distractors that may share some common visual attributes. To address this problem, we present an approach to estimate image saliency by measuring the joint visual surprise from intrinsic and extrinsic contexts. In this approach, a hierarchical context model is first built on a database of 31.2 million images, where a Gaussian mixture model (GMM) is trained for each leaf node to encode the prior knowledge on “what is where” in a specific scene. For a testing image that shares similar spatial layout within a scene, the pre-trained GMM can serve as an extrinsic context model to measure the “surprise” of an image patch. Since human attention may quickly shift between different surprising locations, we adopt a Markov chain to model a surprise-driven attention-shifting process so as to infer the salient patches that can best capture human attention. Experiments show that our approach outperforms 19 state-of-the-art methods in fixation prediction. © 2016 Springer Science+Business Media New York

Li B.,Dalian University of Technology | Zhang Y.,Dalian University of Technology | Lin Z.,Peking University | Lin Z.,Cooperative Medianet Innovation Center | Lu H.,Dalian University of Technology
Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition | Year: 2015

Subspace clustering is a problem of finding a multi-subspace representation that best fits sample points drawn from a high-dimensional space. The existing clustering models generally adopt different norms to describe noise, which is equivalent to assuming that the data are corrupted by specific types of noise. In practice, however, noise is much more complex. So it is inappropriate to simply use a certain norm to model noise. Therefore, we propose Mixture of Gaussian Regression (MoG Regression) for subspace clustering by modeling noise as a Mixture of Gaussians (MoG). The MoG Regression provides an effective way to model a much broader range of noise distributions. As a result, the obtained affinity matrix is better at characterizing the structure of data in real applications. Experimental results on multiple datasets demonstrate that MoG Regression significantly outperforms state-of-the-art subspace clustering methods. © 2015 IEEE.

Li Z.,Peking University | Li Z.,Algorithm | Zhao D.,Algorithm | Lin Z.,Peking University | And 2 more authors.
Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition | Year: 2015

The Riemannian three-factor matrix completion (R3MC) algorithm is one of the state-of-the-art geometric optimization methods for the low-rank matrix completion problem. It is a nonlinear conjugate-gradient method optimizing on a quotient Riemannian manifold. In the line search step, R3MC approximates the minimum point on the searching curve by minimizing on the line tangent to the curve. However, finding the exact minimum point by iteration is too expensive. We address this issue by proposing a new retraction with a minimizing property. This special property provides the exact minimization for the line search by establishing correspondences between points on the searching curve and points on the tangent line. Accelerated R3MC, which is R3MC equipped with this new retraction, outperforms the original algorithm and other geometric algorithms for matrix completion in our empirical study. © 2015 IEEE.

Zhang Z.,Shanghai JiaoTong University | Chen Z.,Shanghai JiaoTong University | Chen Z.,Cooperative Medianet Innovation Center | Shen M.,Shanghai JiaoTong University | Xia B.,Shanghai JiaoTong University
IEEE Journal on Selected Areas in Communications | Year: 2016

In this paper, we consider a multipair amplify-and-forward two-way relay channel, where multiple pairs of full-duplex users exchange information through a full-duplex relay with massive antennas. For improving the energy efficiency, four typical power-scaling schemes are proposed based on the maximum-ratio combining/maximum-ratio transmission (MRC/MRT) and zero-forcing reception/zero-forcing transmission (ZFR/ZFT) at the relay. When the number of relay antennas tends to infinity, we quantify the asymptotic spectral efficiencies and energy efficiencies for the proposed power-scaling schemes. We show that the loop interference can be reduced by decreasing the transmit power under massive relay antennas. Besides, the inter-pair interference and inter-user interference in such systems can also be eliminated in large number of antennas. Moreover, we analytically compare the performance between MRC/MRT and ZFR/ZFT, and describe the impact of the number of user pairs on the spectral efficiency. We also evaluate the energy efficiency performance based on the practical power consumption model, and depict the impact of the relay antenna number on the energy efficiencies for the proposed schemes. Furthermore, we provide the available regions where full-duplex systems can outperform half-duplex systems. Finally, we show that the proposed schemes achieve good performance tradeoffs between the spectral efficiency and the energy efficiency. © 2016 IEEE.

Chen E.,Shanghai JiaoTong University | Tao M.,Shanghai JiaoTong University | Tao M.,Cooperative Medianet Innovation Center
2015 IEEE/CIC International Conference on Communications in China, ICCC 2015 | Year: 2015

Cloud radio access network (Cloud RAN) is able to significantly improve the network capacity and energy efficiency through centralized processing of multiple distributed base stations (BSs). Yet it also places tremendous burden on backhaul links that connect these BSs. In this paper, we introduce a cache-enabled Cloud RAN, in which each BS is equipped with a local cache with limited storage capacity. If the content requested by a user is not cached at the serving BSs of the user, it will be distributed to the serving BSs via the backhaul links from the central processor (CP). We investigate the dynamic user-centric BS clustering and sparse beamforming by taking both channel condition and cache status into account. We formulate an optimization problem with the objective of minimizing the weighted sum of backhaul cost and transmit power cost subject to an individual signal-to-interference-and-noise ratio (SINR) constraint for each user. This problem is a mixed-integer nonlinear programming (MINLP) problem. We apply the iterative reweighted ℓ1-norm technique to find an approximate solution. Theoretical analysis also shows that all the BSs which cache the content requested by a user can always be included in the BS cluster of this user, regardless of their channel conditions. Based on this finding, we propose a cache-aware greedy selection BS clustering algorithm. Simulation results show that caching can greatly improve the tradeoff between backhaul capacity and transmit power. It is also demonstrated that popularity-aware caching is superior to random caching in Cloud RAN. © 2015 IEEE.

Wang Z.,Shanghai JiaoTong University | Chen Z.,Shanghai JiaoTong University | Chen Z.,Cooperative Medianet Innovation Center | Xia B.,Shanghai JiaoTong University | And 2 more authors.
IEEE Transactions on Wireless Communications | Year: 2016

In this paper, a wireless energy harvesting and information transfer protocol in cognitive relay networks is investigated, where an energy harvesting secondary network shares the spectrum as well as harvests energy by assisting the primary transmission. Specifically, a secondary transmitter scavenges energy from the received primary signal and then employs the harvested energy to forward the resulting signals along with the secondary signal. The secondary receiver can also harvest the ambient energy, and use the remaining signal to cancel the primary interference. We analytically derive the exact expressions of the outage probabilities for both primary and secondary networks. The rate-energy tradeoff between the ergodic capacity and harvested energy in the secondary network is also discussed. Furthermore, to quantify the energy consumption, we investigate the system energy efficiency. Moreover, we address the optimization power allocation strategy under three performance criteria and theoretically prove that the resulting nonconvex optimization problems can be converted into biconvex problems. The corresponding effective algorithms are then developed to solve the optimization problems. Numerical results show that the proposed protocol not only achieves both the primary and secondary transmissions but also harvests the ambient energy. © 2015 IEEE.

Zhang Y.,Northwest University, China | Liu J.,Beijing Institute of Technology | Yang W.,Beijing Institute of Technology | Guo Z.,Beijing Institute of Technology | Guo Z.,Cooperative Medianet Innovation Center
IEEE Transactions on Image Processing | Year: 2015

Sparse representation has recently attracted enormous interests in the field of image restoration. The conventional sparsity-based methods enforce sparse coding on small image patches with certain constraints. However, they neglected the characteristics of image structures both within the same scale and across the different scales for the image sparse representation. This drawback limits the modeling capability of sparsity-based super-resolution methods, especially for the recovery of the observed low-resolution images. In this paper, we propose a joint super-resolution framework of structure-modulated sparse representations to improve the performance of sparsity-based image super-resolution. The proposed algorithm formulates the constrained optimization problem for high-resolution image recovery. The multistep magnification scheme with the ridge regression is first used to exploit the multiscale redundancy for the initial estimation of the high-resolution image. Then, the gradient histogram preservation is incorporated as a regularization term in sparse modeling of the image super-resolution problem. Finally, the numerical solution is provided to solve the super-resolution problem of model parameter estimation and sparse representation. Extensive experiments on image super-resolution are carried out to validate the generality, effectiveness, and robustness of the proposed algorithm. Experimental results demonstrate that our proposed algorithm, which can recover more fine structures and details from an input low-resolution image, outperforms the state-of-the-art methods both subjectively and objectively in most cases. © 1992-2012 IEEE.

Peng P.,Beijing University of Technology | Peng P.,Cooperative Medianet Innovation Center | Xiang T.,Queen Mary, University of London | Wang Y.,Beijing Institute of Technology | And 5 more authors.
Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition | Year: 2016

Most existing person re-identification (Re-ID) approaches follow a supervised learning framework, in which a large number of labelled matching pairs are required for training. This severely limits their scalability in realworld applications. To overcome this limitation, we develop a novel cross-dataset transfer learning approach to learn a discriminative representation. It is unsupervised in the sense that the target dataset is completely unlabelled. Specifically, we present an multi-task dictionary learning method which is able to learn a dataset-shared but targetdata-biased representation. Experimental results on five benchmark datasets demonstrate that the method significantly outperforms the state-of-the-art.

Liu H.,Beijing University of Technology | Liu H.,Cooperative Medianet Innovation Center | Tian Y.,Beijing University of Technology | Tian Y.,Cooperative Medianet Innovation Center | And 5 more authors.
Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition | Year: 2016

The growing explosion in the use of surveillance cameras in public security highlights the importance of vehicle search from a large-scale image or video database. However, compared with person re-identification or face recognition, vehicle search problem has long been neglected by researchers in vision community. This paper focuses on an interesting but challenging problem, vehicle re-identification (a.k.a precise vehicle search). We propose a Deep Relative Distance Learning (DRDL) method which exploits a two-branch deep convolutional network to project raw vehicle images into an Euclidean space where distance can be directly used to measure the similarity of arbitrary two vehicles. To further facilitate the future research on this problem, we also present a carefully-organized largescale image database "VehicleID", which includes multiple images of the same vehicle captured by different realworld cameras in a city. We evaluate our DRDL method on our VehicleID dataset and another recently-released vehicle model classification dataset "CompCars" in three sets of experiments: vehicle re-identification, vehicle model verification and vehicle retrieval. Experimental results show that our method can achieve promising results and outperforms several state-of-the-art approaches.

Loading Cooperative Medianet Innovation Center collaborators
Loading Cooperative Medianet Innovation Center collaborators