A Performance Evaluation of Classic Convolutional Neural Networks for 2D and 3D Palmprint and Palm Vein Recognition

Palmprint recognition and palm vein recognition are two emerging biometrics technologies. In the past two decades, many traditional methods have been proposed for palmprint recognition and palm vein recognition, and have achieved impressive results. However, the research on deep learning-based palmprint recognition and palm vein recognition is still very preliminary. In this paper, in order to investigate the problem of deep learning based 2D and 3D palmprint recognition and palm vein recognition in-depth, we conduct performance evaluation of seventeen representative and classic convolutional neural networks (CNNs) on one 3D palmprint database, five 2D palmprint databases and two palm vein databases. A lot of experiments have been carried out in the conditions of different network structures, different learning rates, and different numbers of network layers. We have also conducted experiments on both separate data mode and mixed data mode. Experimental results show that these classic CNNs can achieve promising recognition results, and the recognition performance of recently proposed CNNs is better. Particularly, among classic CNNs, one of the recently proposed classic CNNs, i.e., EfficientNet achieves the best recognition accuracy. However, the recognition performance of classic CNNs is still slightly worse than that of some traditional recognition methods.

     

基于国产处理器增量式实时交通流预测算法及实现

针对城市交通难以处理大量数据且实时性差等问题,提出了根据增量式城市交通流数据预测拥堵情况的一种基于国产处理器的L-BFGS(limited-memory BFGS)算法。该算法通过存储向量序列计算Hessian矩阵,改进Two-Loop算法求下降方向,在Spark集群中并行处理时收敛速度快,适用于实时性要求强的城市交通场景。实验结果证明,L-BFGS预测算法完全可以在国产平台上对大规模的实时交通数据流进行快速建模、预测,在改善城市交通管理水平提供有效支撑的同时也丰富了国产芯片的应用领域。     

认知任务驱动的fMRI脑功能网络构建方法

针对脑功能网络的构建受到特定大脑图谱对兴趣点描述准确度及覆盖度的限制,提出了基于认知任务信息和神经影像数据的脑功能网络构建方法。首先计算fMRI体素对于认知任务的敏感度,然后在此基础上选取兼顾分布均衡和去中心化的大脑兴趣点,从而构建任务驱动的脑功能网络。实验通过在人脸情绪识别认知任务相关的梭状回构造任务驱动的脑功能网络,其度中心性、聚类系数、全局效率、局部效率这四个复杂网络指标均优于典型大脑图谱梭状回中兴趣点构成的网络。结果表明,计算得到的大脑兴趣点具有更强的整合性,更适合用于表征特定认知任务下的脑功能网络。     

基于特征变换的图像检索对抗防御

对抗攻击在图像分类中较早被研究,目的是产生可以误导神经网络预测的不可察觉的扰动.最近,图像检索中的对抗攻击也被广泛探索,研究结果表明最先进的基于深度神经网络的图像检索模型同样容易受到干扰,从而将不相关的图像返回.文中首次尝试研究无需训练的图像检索模型的对抗防御方法,根据图像基本特征因素对输入图像进行变换,以在预测阶段消除对抗攻击的影响.所提方法探索了4种图像特征变换方案,即调整大小、填充、总方差最小化和图像拼接,这些都是在查询图像被送入检索模型之前对其执行的.文中提出的防御方法具有以下优点:1)不需要微调和增量训练过程;2)仅需极少的额外计算;3)多个方案可以灵活集成.大量实验的结果表明,提出的变换策略在防御现有的针对主流图像检索模型的对抗攻击方面是非常有效的.     

基于质心法的车联网目标跟踪方法与应用

车辆目标跟踪是实现车联网不可或缺的一环,旨在获取车辆的动态信息,以提高交通运行效率.其核心是对大量监控探头采集的视频图像进行分析处理,实现车辆的实时检测与跟踪.为了进一步提高目标检测效率,降低硬件成本,文中提出了基于二帧差分法的前景检测方法,以及基于质心法的车辆轮廓检测与跟踪方法.基于OpenCV3.4.1和VS2017进行验证实验和仿真测试,结果表明,该算法对车辆跟踪的精确率达到89.1％,平均处理耗时42.63 ms,具有较好的实时性和鲁棒性,可在车联网嵌入式设备上进行部署和应用.     

基于用户兴趣词典和LSTM的个性化情感分类方法

微博是一个可以分享生活、发表看法、发泄情感的社交平台,由于数据量大且易于获取,微博数据已被广泛用于网络用户情感分析.传统对微博进行情感预测的研究没有考虑用户的用词喜好、语言风格等个性化因素的影响,使得情感分类结果的准确性不高.首先通过分析用户兴趣特征构建用户兴趣词典,在此基础上提出基于用户兴趣词典的情感分类模型;然后利用长短期记忆网络(Long Short-Term Memory,LSTM)分类准确性高的特点训练一个通用的LSTM分类模型;最后利用支持向量机融合不同模型以得到最终的情感分类结果.实验结果表明,与支持向量机、朴素贝叶斯等传统分类器相比,基于用户兴趣词典与LSTM的个性化情感分类方法在分类精度上有较大提升;与LSTM、循环神经网络等深度学习方法相比,该方法在保证运行效率的前提下能获得更高的分类精度.     

An Adaptive Fault-tolerant Control Method for Robot Manipulators

International Journal of Control, Automation and Systems - This paper presents a new method for adaptive continuous nonsingular fast terminal sliding mode control (ACNFTSMC) based on a novel...     

Expectation Maximization Algorithm for Time-delay Output-error Models Based on Finite Impulse Response Method

International Journal of Control, Automation and Systems - In this paper, an output error (OE) model with random time delay is identified by using the expectation maximization (EM) algorithm. Since...     

Power Smoothing of a Variable-speed Wind Turbine Generator

This paper presents a power-smoothing scheme of a variable-speed wind turbine generator (VSWTG) that employs separate control gains for the over-frequency section (OFS) and under-frequency section (UFS). In the proposed scheme, an additional proportional control loop based on the system frequency deviation operating in conjunction with maximum power point tracking operation is used. In the OFS, to improve the energy-storing capability, the scheme suggests the gain of the frequency deviation control loop, which is set to be monotonously decreasing with the rotor speed while being significantly larger than that in the UFS. In the UFS, to improve the energy-releasing capability while preventing over-deceleration, the gain of the frequency deviation control loop is set to be a linear function of the rotor speed. The simulation results under continuously varying wind speeds with different wind patterns and wind speeds clearly demonstrate that the proposed scheme significantly mitigates the output power fluctuations of a VSWTG. The proposed scheme keeps the frequency within a narrow range, thereby reducing the required primary frequency control reserve for regulating the frequency under normal operations.