Time filter

Source Type

Sun B.,Hunan University | Li S.,Hunan University | Sun J.,Fujitsu Research and Development Center
IEEE Transactions on Image Processing | Year: 2014

Currently, most electrophotographic printers use halftoning technique to print continuous tone images, so scanned images obtained from such hard copies are usually corrupted by screen like artifacts. In this paper, a new model of scanned halftone image is proposed to consider both printing distortions and halftone patterns. Based on this model, an adaptive filtering based descreening method is proposed to recover high quality contone images from the scanned images. Image redundancy based denoising algorithm is first adopted to reduce printing noise and attenuate distortions. Then, screen frequency of the scanned image and local gradient features are used for adaptive filtering. Basic contone estimate is obtained by filtering the denoised scanned image with an anisotropic Gaussian kernel, whose parameters are automatically adjusted with the screen frequency and local gradient information. Finally, an edge-preserving filter is used to further enhance the sharpness of edges to recover a high quality contone image. Experiments on real scanned images demonstrate that the proposed method can recover high quality contone images from the scanned images. Compared with the state-of-the-art methods, the proposed method produces very sharp edges and much cleaner smooth regions. © 2014 IEEE. Source

Wu J.,Fujitsu Research and Development Center
IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC | Year: 2012

In this paper, we newly design a CSI-RS for virtualized antenna system when LTE and LTE-A are efficiently coexisting in Rel-10 and beyond. To this end, we newly insert a small amount of CSI-RS into the locations which is close to the existing CRS positions for legacy LTE as enough as possible. The newly designed CSI-RS modulated by space time coded signal efficiently works well for LTE-A UE channel estimation when the number of antennas involved in LTE-A system is larger than four which are normally used for LTE system as maximal. In such a legacy system with eight physical transmit antennas, eight antenna ports with eight orthogonal CSI-RSs are exploited for LTE-A system while virtualized four antenna ports are for LTE system. The benefit of the proposed solution is to significantly reduce the CSI-RS overhead, compared to the overhead employed in the existing Rel-10 system. © 2012 IEEE. Source

Shi Z.,Harbin Institute of Technology | Shi Z.,Fujitsu Research and Development Center | Han J.,Harbin Institute of Technology | Zheng T.,Harbin Institute of Technology | Li J.,A+ Network
IJCAI International Joint Conference on Artificial Intelligence | Year: 2013

This paper studies the recovery guarantees of the models of minimizing ∥Χ∥* + 1/2α ∥Χ∥ F 2 where Χ is a tensor and ∥Χ∥* and ∥Χ∥ F are the trace and Frobenius norm of respectively. We show that they can efficiently recover low-rank tensors. In particular, they enjoy exact guarantees similar to those known for minimizing ∥Χ∥* under the conditions on the sensing operator such as its null-space property, restricted isometry property, or spherical section property. To recover a low-rank tensor Χ0, minimizing ∥Χ∥* + 1/2α ∥Χ∥ F 2 returns the same solution as minimizing ∥Χ∥* almost whenever α ≥ 10 imax ∥X(i) 0∥2. Source

Wang P.,Beihang University | Wang P.,Fujitsu Research and Development Center | Shen C.,University of Adelaide | Barnes N.,NICTA | Zheng H.,Beihang University
IEEE Transactions on Neural Networks and Learning Systems | Year: 2012

Boosting-based object detection has received significant attention recently. In this paper, we propose totally corrective asymmetric boosting algorithms for real-time object detection. Our algorithms differ from Viola and Jones' detection framework in two ways. Firstly, our boosting algorithms explicitly optimize asymmetric loss of objectives, while AdaBoost used by Viola and Jones optimizes a symmetric loss. Secondly, by carefully deriving the Lagrange duals of the optimization problems, we design more efficient boosting in that the coefficients of the selected weak classifiers are updated in a totally corrective fashion, in contrast to the stagewise optimization commonly used by most boosting algorithms. Column generation is employed to solve the proposed optimization problems. Unlike conventional boosting, the proposed boosting algorithms are able to de-select those irrelevant weak classifiers in the ensemble while training a classification cascade. This results in improved detection performance as well as fewer weak classifiers in the learned strong classifier. Compared with AsymBoost of Viola and Jones, our proposed asymmetric boosting is nonheuristic and the training procedure is much simpler. Experiments on face and pedestrian detection demonstrate that our methods have superior detection performance than some of the state-of-the-art object detectors. © 2012 IEEE. Source

Wu J.,Fujitsu Research and Development Center
IEEE International Conference on Communications | Year: 2010

In this paper, we newly design a non-uniform QAM signal constellation, which could be appropriately applied for hierarchical modulation. This novel constellation is constructed by a certain number of equilateral triangles and hierarchically divided into two layers; one for MSB and the other for LSB. The advantage of using this constellation compared to uniform rectangular QAM constellation is that the MED for LSB is constantly maintained whereas the MED for MSB is remarkably increased. To this end, we evaluate the performance in terms of SER by means of link level simulation. Consequently, the newly designed constellation does not compromise the modulation performance between MSB and LSB, and the SNR gain of MSB as opposed to uniform rectangular QAM constellation reaches 0.5∼0.8dB but without sacrificing the performance of LSB. ©2010 IEEE. Source

Discover hidden collaborations