马尔可夫链蒙特卡洛重要度采样与多目标跟踪

针对强杂波环境下的多目标跟踪问题,提出一种基于马尔可夫链蒙特卡洛重要度采样的跟踪方法．通过马尔可夫链蒙特卡洛实现对联合关联事件的采样,据此计算目标可关联量测数据的边缘关联概率．在联合关联事件求解中利用单目标量测的概率密度进行重要度采样,提高采样效率．马尔可夫链蒙特卡洛重要度采样方法克服了联合概率数据关联中的“组合爆炸”问题,能够在强杂波干扰下较好地实现多目标实时跟踪．通过仿真实验对比分析了算法的跟踪精度和处理的时效性,验证了方法的有效性．     

几种典型的微弱点状多运动目标跟踪算法对比研究

重点研究了序列图像情况下几种典型的微弱点状多运动目标实时跟踪算法,虽然它们都能够完成不同背景环境下目标的全程跟踪,但跟踪性能存在较大的差异,PDA算法具有较高的实时性,但容易出现目标的偏移和聚合现象;JPDA算法理论上解决了多目标数据关联问题,但跟踪过程存在较大误差且由于计算量大难以在工程中应用;基于最大熵高斯聚类算法对模糊隶属度进行了修正,数据关联性高且有效避免了目标的误跟和丢失现象。通过对几种典型算法的仿真分析,为多目标跟踪算法的优化提供可靠依据。     

一种改进的LT码置信传播译码

在删除信道上,短信息字长度的LT码采用置信传播译码算法性能较差。提出了一种改进的置信传播译码算法,此算法在置信传播译码失败时只需运用高斯消元法确定少数猜测比特就可实现成功译码。仿真结果显示,相比于置信传播译码,改进的置信传播译码算法在增加少量译码运行时间的代价下获得较大的译码性能改善。     

Access point selection in heterogeneous wireless networks using belief propagation

The next generation wireless network will be composed by various heterogenous wireless access networks,such as cellular network,worldwide interoperability for microwave access(WiMAX),wireless local area network(WLAN),etc.Different access networks cooperatively provide high-bandwidth connectivity with bandwidth guarantees.This paper proposes a utility-based access point selection scheme,which selects an accessible point for each user,such that the bandwidth requirement of each user is satisfied,and also the defined utility function is maximized.Due to the NP-complete nature of the problem,the existing proposals apply the greedy method to find a solution.We find that belief propagation is an efficient tool to solve this problem,and thus,we derive the same optimization objective in a new way,and then draw a factor graph representation which describes our combinatorial optimization problem.Afterwards,we develop the belief propagation algorithm,and show that our algorithm converges.Finally,we conduct numerical experiments to evaluate the convergency and accuracy of the belief propagation in load balancing problem.     

基于信念重用的WSNs能量高效跟踪

针对无线传感器网络(WSNs)中目标跟踪性能与传感器能量消耗难以平衡问题,提出一种信念重用的WSNs能量高效跟踪算法。使用部分可观察马尔可夫决策过程(POMDPs)对动态不确定环境下的WSNs进行建模,将跟踪性能与能量消耗平衡优化问题转化为POMDPs最优值函数求解过程;采用最大报酬值启发式查找方法获得跟踪性能的逼近最优值;采用信念重用方法避免重复获取信念,有效降低传感器通信带来的能量消耗。实验结果表明:信念重用算法能够有效优化跟踪性能与能量消耗之间的平衡,达到以较低的能量消耗获得较高跟踪性能的目的。     

在癌症分类中基于分层抽样的神经网络集成算法

在基因表达谱数据的分析中,针对有效合理地选择特征基因集的问题,本文将分层抽样技术引入特征基因选择,提高特征基因集的分类能力。以神经网络作为分量分类器,神经网络集成进行分类预测。并在结肠癌数据集上进行实验,实验结果表明该方法能有效地降低特征基因集选择的复杂性,提高对于未知样本的分类预测效果。     

Online belief propagation algorithm for probabilistic latent semantic analysis

Probabilistic latent semantic analysis (PLSA) is a topic model for text documents, which has been widely used in text mining, computer vision, computational biology and so on. For batch PLSA inference algorithms, the required memory size grows linearly with the data size, and handling massive data streams is very difficult. To process big data streams, we propose an online belief propagation (OBP) algorithm based on the improved factor graph representation for PLSA. The factor graph of PLSA facilitates the classic belief propagation (BP) algorithm. Furthermore, OBP splits the data stream into a set of small segments, and uses the estimated parameters of previous segments to calculate the gradient descent of the current segment. Because OBP removes each segment from memory after processing, it is memory-efficient for big data streams. We examine the performance of OBP on four document data sets, and demonstrate that OBP is competitive in both speed and accuracy for online expectation maximization (OEM) in PLSA, and can also give a more accurate topic evolution. Experiments on massive data streams from Baidu further confirm the effectiveness of the OBP algorithm.     

多目标跟踪的混合高斯PHD滤波

为解决目标数未知或随时间变化时的多目标跟踪问题,将多目标状态和观测信息表示为随机集的形式,建立了多目标跟踪的混合高斯概率假设密度（PHD）滤波方法。当目标初始的先验概率密度满足高斯分布的形式时,通过将状态噪声、观测噪声、目标的繁衍、新目标的产生、目标的存活概率和检测概率表示成混合高斯的形式,之后每个时刻的后验概率密度均能表示成混合高斯的形式。线性混合高斯PHD滤波方法将Kalman滤波引入到PHD滤波中,利用混合高斯成分预测和更新随机集的PHD,并估计出目标的状态。实验结果表明,在杂波环境下混合高斯PHD滤波方法可以有效地跟踪目标状态。     

TOA-based passive localization of multiple targets with inaccurate receivers based on belief propagation on factor graph

Location awareness is now becoming a vital requirement for many practical applications. In this paper, we consider passive localization of multiple targets with one transmitter and several receivers based on time of arrival (TOA) measurements. Existing studies assume that positions of receivers are perfectly known. However, in practice, receivers' positions might be inaccurate, which leads to localization error of targets. We propose factor graph (FG)-based belief propagation (BP) algorithms to locate the passive targets and improve the position accuracy of receivers simultaneously. Due to the nonlinearity of the likelihood function, messages on the FG cannot be derived in closed form. We propose both sample-based and parametric methods to solve this problem. In the sample-based BP algorithm, particle swarm optimization is employed to reduce the number of particles required to represent messages. In parametric BP algorithm, the nonlinear terms in messages are linearized, which results in closed-form Gaussian message passing on FG. The Bayesian Cramér–Rao bound (BCRB) for passive targets localization with inaccurate receivers is derived to evaluate the performance of the proposed algorithms. Simulation results show that both the sample-based and parametric BP algorithms outperform the conventional method and attain the proposed BCRB. Receivers' positions can also be improved via the proposed BP algorithms. Although the parametric BP algorithm performs slightly worse than the sample-based BP method, it could be more attractive in practical applications due to the significantly lower computational complexity.     

强跟踪输入估计概率假设密度多机动目标跟踪算法

针对多机动目标跟踪中,目标数目未知及加速度不确定的问题,提出一种强跟踪输入估计（modifiedinputestimation,MIE）概率假设密度多机动目标跟踪算法．在详细分析算法的基础上,通过引入强跟踪多重渐消因子,以不同速率实时调节滤波器各个通道的预测协方差及相应的滤波器增益,从而实现MIE算法对加速度未知或发生人幅度突变的机动目标白适应跟踪能力;并将该算法与概率假设密度滤波算法有效结合,町以较好地跟踪未知数目的多机动目标．仿真结果表明,新算法比传统的多机动目标跟踪算法具有更岛的跟踪精度,且具有较好的实时性．     

低视点下遮挡自适应感知的多目标跟踪算法

目的 针对低视点多目标跟踪场景的遮挡问题,提出一种能够遮挡自适应感知的多目标跟踪算法。方法 首先根据每帧图像的全局遮挡状态,提出了“自适应抗遮挡特征”,增强目标特征对遮挡的感知和调整能力。同时,采用“级联筛查机制”,减少由遮挡带来的目标特征剧烈变化而认定为“虚新入目标”的错误跟踪现象。最后,考虑到历史模板库中存在遮挡的模板对跟踪性能的影响,根据每一帧中目标的局部遮挡状态,提出自适应干扰模板更新机制,进一步提高对遮挡的应变和适应能力。结果 实验结果表明,本文算法在MOTA(multiple object tracking accuracy)、M OTP (multiple object tracking precision)、FN(false negatives)、Rcll (recall)、ML (mostly lost tracklets)等指标上明显优于STAM(spatial-temporal attention mechanism)、ATAF(aggregate tracklet appearance features)、STRN (spatial-temporal relat...     

Elastic neural network method for multi-target tracking task allocation in wireless sensor network

Aiming at the task allocation of collaborative technique in wireless sensor network, a method for optimized task allocation based on elastic neural network is proposed under the background of multi-sensor tracking. First a model of multi-coalition tracking multi-target is designed. Then disjoint fully connected subgraphs of neurons are constructed to solve the problem of optimized task allocation in tracking multi-target and the increment of system energy consumption when dynamic coalitions compete and conflict for the resource of sensor nodes. Compared with the conventional method, simulation results show that the energy consumption of the tracking system is reduced significantly and the tracking accuracy is improved greatly, demonstrating the effectiveness of elastic neural network in handling the optimized task allocation problem of multi-sensor tracking multi-target.     

Fuzzy spatial constraints and ranked partitioned sampling approach for multiple object tracking

While particle filters are now widely used for object tracking in videos, the case of multiple object tracking still raises a number of issues. Among them, a first, and very important, problem concerns the exponential increase of the number of particles with the number of objects to be tracked, that can make some practical applications intractable. To achieve good tracking performances, we propose to use a Partitioned Sampling method in the estimation process with an additional feature about the ordering sequence in which the objects are processed. We call it Ranked Partitioned Sampling, where the optimal order in which objects should be processed and tracked is estimated jointly with the object state. Another essential point concerns the modeling of possible interactions between objects. As another contribution, we propose to represent these interactions within a formal framework relying on fuzzy sets theory. This allows us to easily model spatial constraints between objects, in a general and formal way. The association of these two contributions was tested on typical videos exhibiting difficult situations such as partial or total occlusions, and appearance or disappearance of objects. We show the benefit of using conjointly these two contributions, in comparison to classical approaches, through multiple object tracking and articulated object tracking experiments on real video sequences. The results show that our approach provides less tracking errors than those obtained with the classical Partitioned Sampling method, without the need for increasing the number of particles.     

基于箱粒子的多扩展目标PHD滤波

在高斯混合多扩展目标PHD 滤波的基础上, 结合最新兴起的箱粒子滤波, 提出一种基于区间分析的多扩展目标PHD 滤波算法. 采用大小可控的非零矩形区域来代替传统的多个点量测, 这样可降低权值计算中对量测分布的要求. 仿真对比实验表明, 采用区间分析方法在保证近似于传统滤波精度的同时可降低计算复杂度, 在目标数目估计及抗杂波干扰方面也具有较为突出的优势, 并且可解决在目标靠近时由于不能正确给出子划分而造成的漏检问题.

     

基于网络流跟踪的信号灯检测方法

结合信号灯信息对机动车行进速度进行引导,减少机动车启停次数,可有效减少废气排放,缓解其造成的污染问题。针对信号灯转换时刻的获取问题,提出了一种基于网络流跟踪的信号灯检测方法。首先,该方法在数据集中引入辅助信号灯类别进行训练,将视频序列中该类目标检测结果关联为踪片,并通过踪片建模多目标跟踪任务。其次,该方法将多目标跟踪任务转换为最小费用流优化任务,以踪片作为节点建立最小费用流网络,提出了适合于信号灯的费用构建方式,通过最短路径算法求解,得到视频序列中辅助信号灯的多条轨迹。最后,基于求解的轨迹结果和图像分类技术,实现信号灯检测性能的提升。该方法的跟踪性能相较于对比算法有大幅提升,并将小目标信号灯检测响应的mAP提升至94.35%。实验结果表明,基于网络流的建模方式能极大地提升信号灯的跟踪准确率,结合跟踪轨迹还能大幅提高视频序列中小目标信号灯的检测准确率,并可有效确定信号灯状态的转换时刻。     

Modeling temporal processes via belief networks and Petri nets,with application to expert systems

We show how Bayesian belief networks (BNs) can be used to model common temporal knowledge. Two approaches to their structuring are proposed. The first leads to BNs with nodes representing states of a process and times spent in such states, and with a graphical structure reflecting the conditional independence assumptions of a Markovian process. A second approach leads to BNs whose topology represents a conditional independence structure between event-times. Once required distributional specifications are stored within the nodes of a BN, this becomes a powerful inference machine capable, for example, of reasoning backwards in time. We discuss computational difficulties associated with propagation algorithms necessary to perform these inferences, and the reasons why we chose to adopt Monte Carlo-based propagation algorithms. Two improvements to existing Monte Carlo algorithms are proposed; an enhancement based on the principle of importance sampling, and a combined technique that exploits both forward and Markov sampling. Finally, we consider Petri nets, a very interesting and general representation of temporal knowledge. A combined approach is proposed, in which the user structures temporal knowledge in Petri net formalism. The obtained Petri net is then automatically translated into an equivalent BN for probability propagation. Inferred conclusions may finally be explained with the aid of Petri nets again.     

无人机多机动目标自主探测与跟踪

无人机机载相机图像中机动目标尺寸较小而且会发生显著变化,加上大量的背景噪声干扰,给目标探测和跟踪带来很大困难.针对这些问题,本文提出了一种在无人机机载相机图像序列中自主探测与跟踪多个机动目标的方法.首先,提取目标的图像数字特征并采用级联分类算法进行特征分类,得到目标的强分类器,对目标进行自主探测搜索.然后,基于全局最优关联算法对探测回波进行关联滤波,实现对多个机动目标的跟踪与识别,其中最优关联代价矩阵融合了距离和方向信息,提高了关联和跟踪的鲁棒性.将无人机航拍图像序列中的地面坦克作为目标进行实验,结果表明本文算法可以实现对多个机动目标的自主探测和跟踪,并具有较好的跟踪鲁棒性.     

未知噪声统计下多模型概率假设密度粒子滤波算法

针对传统多目标概率假设密度滤波(PHD) 器在噪声先验统计未知或不准确时滤波精度下降甚至丢失目标的问题, 设计一种自适应多模型粒子PHD(MMPHD) 滤波算法. 该算法利用多模型近似思想, 推导出一种多模型概率假设密度估计器, 不仅能估计多目标状态, 而且能实时估计未知且时变的噪声参数, 并采用蒙特卡罗方法给出了MMPHD闭集解. 仿真实例表明, 所提出的算法具有应对噪声变化的自适应能力, 可有效提高目标跟踪精度.

     

Approximate stochastic dynamic programming for sensor scheduling to track multiple targets

The problem of sensor scheduling is to select the number and combination of sensors to activate over time. The goal is usually to trade off tracking performance and sensor usage. We formulate a version of this problem involving multiple targets as a partially observable Markov decision process, and use this formulation to develop a nonmyopic sensor-scheduling scheme. Our scheme integrates sequential multisensor joint probabilistic data association and particle filtering for belief-state estimation, and use a simulation-based Q-value approximation method called completely observable rollout for decision making. We illustrate the effectiveness of our approach by an example with multiple sensors activated simultaneously to track multiple targets. We also explore the trade-off between tracking error and sensor cost using our nonmyopic scheme.     

Multiple importance unscented Kalman filtering with soft spatiotemporal constraint for multi-passive-sensor target tracking

Multi-passive-sensor systems are a common means for the target tracking and their bearings processing is a prerequisite for stable control and nonlinear filtering. This study proposes a mathematical methodology that is based on the incorporating deterministic unscented transition rules into stochastic sequential importance sampling frame and makes use of soft spatiotemporal constraint comprise multiview epipolar geometry constraint and numerical regularization to solve the correspondence problem. A prototype measurement-driven target tracking frame was developed that can work in real time and achieve filtering improvements of 41%–46% and 43%–48% in terms of root-mean-square error and root time-averaged mean square error compared with the state-of-the-art multiple model Rao–Blackwell particle filtering method, as proven by the simulation results.