无线感知执行网的智能协作机制研究

针对无线感知执行网中的节点协作问题,提出了多目标智能协作策略。在感知节点不完全连通的WSANs中,设计局部感知节点成簇算法,实现感知节点间的通信协作;以网络能耗均衡和信息传输时延为目标,应用离散多目标自由搜索算法对WSANs进行实时路径规划,实现簇头节点和执行节点间的协作,并通过路径惩罚策略和路径均衡策略,保证寻优结果的有效性、均衡执行节点之间的能耗。仿真实验表明:无线感知执行网智能协作机制能有效地选择多执行节点路径,实现网络协作的实时需求与能量消耗之间的平衡。     

基于横纵向联合控制的多目标优化车辆跟驰研究

为解决车辆在拥堵环境中因车速波动较大所带来的跟驰平稳性较差、跟踪无效或不安全等问题,提出了基于车辆模型和深度强化学习的多目标优化跟驰方案。首先基于车辆横纵向动力学建立车辆跟驰模型,然后根据车间距误差、速度误差、横向偏差及相对偏航角等,利用深度确定性策略梯度算法得到跟驰车的加速度和转向角,以更平稳安全地控制跟驰车辆。经NGSIM公开驾驶数据集进行测试与验证,该方案可有效地提升跟驰车辆的稳定、舒适与安全性,对保证交通安全和提升道路通行能力具有重要意义。     

自动化学科面临的挑战

本文分析了控制理论与应用、模式识别与智能系统、导航制导与控制、系统科学与工程、人工智能与自动化交叉等领域的发展现状. 结合科技发展、国内国际研究前沿和新兴领域对自动化科学技术的需求, 提出重点发展智能控制理论和方法、高性能作业机器人、信息物理系统、导航与控制技术、重大装备自动化技术、自主智能系统和人工智能驱动的自动化技术优先领域, 加强数据驱动控制理论、人工智能基础理论研究, 进一步发展人机协同、跨域融合的智能自动化, 为实现国家社会的全面信息化智能化提供理论和技术保障.     

Measuring the likelihood of VR visual fatigue through ocular biomechanics

Immersion in virtual reality is still linked to symptoms of visual fatigue such as eye strain, dizziness, and overall discomfort. Studies have investigated visual fatigue through pre- and post-immersion tests of the visual function. In this work, we extend on our previous study and derive a visual fatigue likelihood metric using biomechanical analysis. Previously, we have investigated the effect of VR on the vergence system during immersion. The proposed visual fatigue metric exhibited a significant correlation to vergence angle variability which was previously linked to vergence accommodation conflict in VR. We also discuss subjective feedback and its relationship with the proposed visual fatigue metric.     

基于“雁阵效应” 的扑翼飞行机器人高效集群编队研究

本文借鉴“雁阵效应”, 研究了扑翼飞行机器人高效集群编队飞行问题. 通过对“V”字雁阵的分析得知, 当前排大雁(简称头雁)和后排大雁(简称从雁)保持某一合适的相对位置偏移时, 后排大雁可有效利用前排大雁挥翅产生的上洗涡流, 从而节省体能; 并且, 雁阵通过阵型的变换, 可以实现能量整体消耗的均衡性, 确保长航时飞行. 仿照该“雁阵效应”, 分析得出耗能最少的扑翼飞行机器人集群阵型排布方式, 并设计了阵型变换机制, 实现集群能量整体消耗的最优性和均衡性. 在此基础上, 参考雁群的交互方式, 设计了一种使用局部信息的控制方法, 保证最优阵型的稳定维持以及阵型间的灵活变换. 最后, 仿真结果验证了所提理论结果的有效性.     

Speech Enhancement via Residual Dense Generative Adversarial Network

Generative adversarial networks (GANs) are paid more attention to dealing with the end-to-end speech enhancement in recent years. Various GAN-based enhancement methods are presented to improve the quality of reconstructed speech. However, the performance of these GAN-based methods is worse than those of masking-based methods. To tackle this problem, we propose speech enhancement method with a residual dense generative adversarial network (RDGAN) contributing to map the log-power spectrum (LPS) of degraded speech to the clean one. In detail, a residual dense block (RDB) architecture is designed to better estimate the LPS of clean speech, which can extract rich local features of LPS through densely connected convolution layers. Meanwhile, sequential RDB connections are incorporated on various scales of LPS. It significantly increases the feature learning flexibility and robustness in the time-frequency domain. Simulations show that the proposed method achieves attractive speech enhancement performance in various acoustic environments. Specifically, in the untrained acoustic test with limited priors, e.g., unmatched signal-to-noise ratio (SNR) and unmatched noise category, RDGAN can still outperform the existing GAN-based methods and masking-based method in the measures of PESQ and other evaluation indexes. It indicates that our method is more generalized in untrained conditions.     

AdaFuse: Adaptive Multiview Fusion for Accurate Human Pose Estimation in the Wild

International Journal of Computer Vision - Occlusion is probably the biggest challenge for human pose estimation in the wild. Typical solutions often rely on intrusive sensors such as IMUs to...     

一种改进的自适应图像颜色迁移算法

颜色迁移是图像处理中一个重要的研究课题,它要解决问题是:基于图像A和图像B,合成一幅新的图像B,使其具有A的颜色和B的形状;提出了一种改进的自适应图像颜色迁移算法,它可以更好地利用图像的局部信息,并在算法中引入了正交化分析,进一步减少了颜色空间各通道间的相关性,实现对图像A和图像B的图像颜色最优的迁移过程。     

基于Vega的舰载捷联惯导虚拟平台的航迹仿真

运用虚拟现实技术(Virtual Reality;VR)对捷联惯导系统(Strapdown Inertial Navigation System,SINS)进行模拟的方案,可以有效地提高研发效率,节约研发成本;现在利用Multigen Vega交互仿真环境与VC++6.0开发环境,建立虚拟方针环境;随之针对建立的舰载惯性导航系统航迹仿真的数学模型,分别利用毕卡逼近算法、双子样算法和三子样算法,进行导航解算;最后由真值和所得解算值,实时同步绘制了原始轨迹和解算轨迹,实现了在模拟平台下的航迹仿真;两组轨迹的结果显示的误差,能直观显示出算法的计算误差大小;如此,使得惯性系统的仿真数据可以更加直观的表现出来。     

Automatic execution of workflows on laser-scanned data for extracting bridge surveying goals

With the capability of capturing detailed geometry of bridges in minutes, laser scanning technology has attracted the interests of bridge inspectors and researchers in the domain of bridge management. A challenge of effectively utilizing laser scanned point clouds for bridge inspection is that inspectors need to manually extract and measure large numbers of geometric features (e.g., points) for deriving geometric information items (e.g., the minimum underclearance) of bridges, named as bridge surveying goals in this research. Tedious manual data processing impedes inspectors from quantitatively understanding how various data processing options (e.g., algorithms, parameter values) influence the data processing time and the reliabilities of the surveying goal results. This paper shows the needs of automatic workflow executions for extracting surveying goals from laser scanned point clouds, and presents a computational framework for addressing these needs. This computational framework is composed of formal representations of workflows and mechanisms for constructing and executing workflows. Using a prototype system implemented based on this framework, we constructed and quantitatively characterized three workflows for extracting three representative bridge surveying goals, using three metrics of workflow performance defined in this research: exhaustiveness of measurement sampling, reliability of surveying goal results, and time efficiency.