正交投影神经网络

本文在前馈网络的基础上，提出了一种新的网络模型。在该模型中，由于将线性优化技术引入其学习训练过程，从而提高了网络的学习速度。同时，该网络仍保持了神经网络中非线性映射的基本特点。由于这种网络的学习算法是基于正交投影算法，故我们称之为正交投影神经网络（ＯＰＮＮ）。最后，给出了两个利用ＯＰＮＮ网络分别进行异或分类和函数逼近的例子。     

A fast algorithm of bitstream extraction using distortion prediction based on simulated annealing

Scalable video streams can be extracted to meet the bandwidth limitation of different networks and end-users. Bitstream extraction is usually performed at the network proxy or gateway during transmission, where a low computational complexity is always preferred. How to quickly and accurately select the best resolution combination for a video to meet different bandwidth requirements by each user is crucial in bitstream extraction. In this paper a fast algorithm of bitstream extraction for scalable video is proposed. The interlayer dependency between the base quality layer and the first quality layer was used to predict the distortion of higher quality layers. When quality of every layer is available, the proposed method searches for the optimized combination of quality layers based on simulated annealing. Experimental results show that the proposed method provides an optimized performance, which is significantly higher than that can be achieved by the basic extraction method. Compared to the quality layer based extraction method in the reference software model of H.264/SVC (i.e., JSVM), the proposed algorithm can greatly decrease the decoding times from 2NT to only 2 without losing rate-distortion performance. Furthermore, the proposed method obtains a more smoothed video quality which is always favorable to the observer.     

Clean Chromite Production from Fine Chromite Tailings by Combination of Multi Gravity Separator and Hydrocyclone

Chromite is an important mineral used in the metallurgy, chemistry, and refractory industries. For this reason, beneficiation of chromite tailings is very important. In this study, the possibility of beneficiation of chromite tailings in the Uckopru/Fethiye-Turkey region by Hydrocyclone and Multi Gravity Separator (MGS) combination was investigated. The two signification operational parameters of hydrocyclone, which were diameter of apex and diameter of vortex, and the three signification operational parameters of MGS, which were drum speed, tilt angle, and wash water, were varied and the results were evaluated with the Central Composite Rotatable Design. The results of beneficiation studies showed that commercial concentrate containing 48.18% Cr₂O₃ was obtainable with a 69.79% rate of recovery.     

MGS片层重要性分析及其应用*

基于中粒度质量可伸缩（medium grain scalable, MGS）编码在可伸缩编码（scalable video coding, SVC）应用中的广泛性,在分析了MGS编码结构特点和问题的基础上,利用DCT系数分布特性提出MGS片层数据重要性的估计方法实现对MGS编码各片层重要性的估计。在此重要性估计方法的前提基础上,将之应用于指导可伸缩编码中的码流提取和容错编码。实验测试结果表明,MGS片层数据重要性估计方法的准确性和应用有效性,保障了视频码流质量,提高了码流提取效率。     

基于内容感知的视频流媒体渐进式流传输方式

为有效利用传输网络带宽、平衡播出视频质量与用户等待时间之间的矛盾,提出一种基于内容感知的视频流媒体渐进式流传输方式。在视频内容分析和理解的基础上,根据编码关键参数确定视频序列中各帧的重要性,估计帧内片层数据重要性。采用MGS编码,在MobileIP测试平台上进行实验。实验结果表明了该传输方式的正确性和有效性,能够适应网络带宽和质量可伸缩的视频流媒体。     

基于贪婪算法的MGS编码优化方法

提出一种基于贪婪算法的MGS可伸缩编码的码率控制优化方法。该方法以较低的计算复杂度优化MGS编码,根据不同帧类型和质量层类型的统计特性分别建立不同的率失真模型,并基于贪婪算法确定MGS的片层优化分配方案。实验结果表明,该方法在相同码率下明显提高解码图像的质量,平均峰值信噪比可提高0.5 dB。     

下一代光网络中的多粒度光交换技术

论述了多粒度光交换技术(MGS)的特点，通过计算比较端口数，定量分析使用MGS能简化光交叉连接器(OXC)的结构。提出了基于MGS的多粒度分层路南、多粒度信令机制以及链路管理技术。讨论了引入MGS对自动交换光网络(ASON)的生存性和流量工程等系统应用方面的影响。     

A fast mode decision algorithm applied in medium-grain quality scalable video coding

To increase the flexibility of bit stream adaptation in the H.264/Scalable Video Coding, Medium-Grain Scalable Video Coding (MGS), a variation of Coarse-Grain Scalable Video Coding (CGS), is included. The encoding structure of MGS is significantly different from that of CGS, and the encoding procedure of MGS is very complex. It is important to reduce encoding complexity without sacrificing MGS coding performance. In this paper, a fast mode decision algorithm is proposed for the enhancement layer of MGS based on its specific coding structure. First, the candidate modes and the coding sequence of the current macro-block (MB) are predicted based on mode correlations and correlation degrees. Next, early termination strategies, including Direct Mode, inter-layer, layer and spatial, and spatial-only, are proposed based on the coding structure of MGS and correlations. Finally, MBs are encoded in the order predicted with the four proposed early termination strategies to improve the coding speed. Experimental results show that the algorithm can achieve an average time saving of up to 85.44% with strong robustness and negligible loss in coding efficiency.