Body-based interfaces

This research explores different ways to use features of one's own body for interacting with computers. Such "body-based" interfaces may find good uses in wearable computing or virtual reality systems as part of a 3D multi-modal interface in the future, freeing the user from holding interaction devices. Four types of body-based interfaces have been identified: Body-inspired metaphor (BIM); Body-as-interaction-surface (BAIS); Mixed mode (MM); and Object mapping (OM). These four body-based interfaces were applied to a few different applications (and associated tasks) and were tested for their performance and preference. It was generally found that, among the four, the BIM exhibited low error rates, but produced relatively longer task completion times and significant fatigue. The BAIS method had the contrasting character of higher error rates, but shorter task completion times and lower intuitiveness. The OM method exhibited high error rates, longer completion times, and much fatigue. Overall, the MM was superior in terms of both performance and preference as it combined the merits of the above three methods. Thus, it is expected, for applications with many associated tasks, a careful division of tasks among those that have natural semantic links to body parts and those that do not, is necessary to design the most performing body-based interface.     

一种新的混合免疫算法

免疫算法是在免疫系统识别多样性的启发下所设计出的一种新的多峰值函数的寻优算法。文章针对目前人工免疫算法在收敛于全局最优方面存在的缺陷,提出了一种模拟退火免疫算法,引入Boltzmann退火选择算子,利用模拟退火算法可选择性接受恶化解的特点克服传统免疫算法的缺陷,并通过实验证明了该混合免疫算法的优越性能。     

求解多峰函数问题的模拟退火算法

论文提出了一种与演化算法群体搜索技术相结合的模拟退火算法(SimulatedAnnealingAlgorithmbasedonIn-dividuals,以下简记为ISA)。与传统的模拟退火算法(SimulatedAnnealing,以下简记为SA)不同的是它由多个个体在搜索空间里同时搜索,个体与个体之间没有信息交换,在遗传算子的作用下每个个体各自朝着适应值高的方向迅速演化,因而只要初始群体是均匀、随机地分布在搜索空间里并且个体数目适当多,算法最终都能找到多峰函数的所有峰值点。     

一种改进的演化算法

将GT算法和粒子群优化（PSO）算法结合并加以改进，采用non uniform变异算子提高局部搜索能力和算法的稳定性，同时引入种群划分等策略，构建了一种新的演化算法(记为GT POPDM PSO)。该算法比上述两种算法具有更好的性能，特别是对多峰函数优化等问题计算效果更好。     

多媒体交叉参照检索系统研究

描述并实现了一个面向数字图书馆的多媒体交叉参照检索系统．该系统将对多模态数据的检索无缝地融合到一个系统中,基于交叉参照图模型,多模态交叉参照搜索引擎综合计算媒体对象与查询之间在语义和内容层上的相似度;系统还支持用户相关反馈调整交叉参照图模型．实验结果表明,该系统能够比较有效地进行多模态交叉参照检索．     

Neuroinflammation and Precision Medicine in Pediatric Neurocritical Care: Multi-Modal Monitoring of Immunometabolic Dysfunction

The understanding of molecular biology in neurocritical care (NCC) is expanding rapidly and recognizing the important contribution of neuroinflammation, specifically changes in immunometabolism, towards pathological disease processes encountered across all illnesses in the NCC. Additionally, the importance of individualized inflammatory responses has been emphasized, acknowledging that not all individuals have the same mechanisms contributing towards their presentation. By understanding cellular processes that drive disease, we can make better personalized therapy decisions to improve patient outcomes. While the understanding of these cellular processes is evolving, the ability to measure such cellular responses at bedside to make acute care decisions is lacking. In this overview, we review cellular mechanisms involved in pathological neuroinflammation with a focus on immunometabolic dysfunction and review non-invasive bedside tools that have the potential to measure indirect and direct markers of shifts in cellular metabolism related to neuroinflammation. These tools include near-infrared spectroscopy, transcranial doppler, elastography, electroencephalography, magnetic resonance imaging and spectroscopy, and cytokine analysis. Additionally, we review the importance of genetic testing in providing information about unique metabolic profiles to guide individualized interpretation of bedside data. Together in tandem, these modalities have the potential to provide real time information and guide more informed treatment decisions.     

Memory based fusion for multi-modal deep learning

The use of multi-modal data for deep machine learning has shown promise when compared to uni-modal approaches with fusion of multi-modal features resulting in improved performance in several applications. However, most state-of-the-art methods use naive fusion which processes feature streams independently, ignoring possible long-term dependencies within the data during fusion. In this paper, we present a novel Memory based Attentive Fusion layer, which fuses modes by incorporating both the current features and long-term dependencies in the data, thus allowing the model to understand the relative importance of modes over time. We introduce an explicit memory block within the fusion layer which stores features containing long-term dependencies of the fused data. The feature inputs from uni-modal encoders are fused through attentive composition and transformation followed by naive fusion of the resultant memory derived features with layer inputs. Following state-of-the-art methods, we have evaluated the performance and the generalizability of the proposed fusion approach on two different datasets with different modalities. In our experiments, we replace the naive fusion layer in benchmark networks with our proposed layer to enable a fair comparison. Experimental results indicate that the MBAF layer can generalize across different modalities and networks to enhance fusion and improve performance.     

基于注意力机制的多模态人体行为识别算法

提出了基于注意力机制的多模态人体行为识别算法;针对多模态特征的有效融合问题,设计基于注意力机制的双流特征融合卷积网络(TAM3DNet,two-stream attention mechanism 3D network);主干网络采用结合注意力机制的注意力3D网络(AM3DNet,attention mechanism 3D network),将特征图与注意力图进行加权后得到加权行为特征,从而使网络聚焦于肢体运动区域的特征,减弱背景和肢体静止区域的影响;将RGB-D数据的颜色和深度两种模态数据分别作为双流网络的输入,从两条分支网络得到彩色和深度行为特征,然后将融合特征进行分类得到人体行为识别结果。     

Spiral bacterial foraging optimization method: Algorithm,evaluation and convergence analysis

A biologically-inspired algorithm called Spiral Bacterial Foraging Optimization (SBFO) is investigated in this article. SBFO, previously proposed by the same authors, is a multi-agent, gradient-based algorithm that minimizes both the main objective function (local cost) and the distance between each agent and a temporary central point (global cost). A random jump is included normal to the connecting line of each agent to the central point, which produces a vortex around the temporary central point. This random jump is also suitable to cope with premature convergence, which is a feature of swarm-based optimization methods. The most important advantages of this algorithm are as follows: First, this algorithm involves a stochastic type of search with a deterministic convergence. Second, as gradient-based methods are employed, faster convergence is demonstrated over GA, DE, BFO, etc. Third, the algorithm can be implemented in a parallel fashion in order to decentralize large-scale computation. Fourth, the algorithm has a limited number of tunable parameters, and finally SBFO has a strong certainty of convergence which is rare in existing global optimization algorithms. A detailed convergence analysis of SBFO for continuously differentiable objective functions has also been investigated in this article.     

基于介质谐振器的轨道角动量天线设计

近年来,轨道角动量一直是无线通信领域的研究热点。本文将轨道角动量与毫米波技术相结合,提出了一种基于介质谐振器的轨道角动量天线。建立了该天线的等效模型,推导了其辐射场的理论表达,讨论了介质谐振器半径对涡旋波电磁波模态的影响,通过仿真结果表明,该天线在K_(a)波段有四个谐振点,能够分别产生模态l=1,2,3,4的涡旋电磁波。此外,该天线结构紧凑、成本低、增益良好,具有较高的天线效率,产生的各个模态的涡旋电磁波都具有良好的旋转性,能够获得较强的抗干扰能力,为轨道角动量在毫米波频段的应用提供了一定的现实意义。