Recognizing tactic patterns in broadcast basketball video using player trajectory

The explosive growth of the sports fandom inspires much research on manifold sports video analyses and applications. The audience, sports fans, and even professionals require more than traditional highlight extraction or semantic summarization. Computer-assisted sports tactic analysis is inevitably in urging demand. Recognizing tactic patterns in broadcast basketball video is a challenging task due to its complicated scenes, varied camera motion, frequently occlusions between players, etc. In basketball games, the action screen means that an offensive player perform a blocking move via standing beside or behind a defender for freeing a teammate to shoot, to receive a pass, or to drive in for scoring. In this paper, we propose a screen-strategy recognition system capable of detecting and classifying screen patterns in basketball video. The proposed system automatically detects the court lines for camera calibration, tracks players, and discriminates the offensive/defensive team. Player trajectories are calibrated to the real-world court model for screen pattern recognition. Our experiments on broadcast basketball videos show promising results. Furthermore, the extracted player trajectories and the recognized screen patterns visualized on a court model indeed assist the coach/players or the fans in comprehending the tactics executed in basketball games informatively and efficiently.     

CDADNet: Context-guided dense attentional dilated network for crowd counting

Crowd counting is a conspicuous task in computer vision owing to scale variations, perspective distortions, and complex backgrounds. Existing research usually adopts the dilated convolution network to enlarge the receptive fields to solve the problem of scale variations. However, these methods easily bring background information into the large receptive fields to generate poor quality density maps. To address this problem, we propose a novel backbone called Context-guided Dense Attentional Dilated Network (CDADNet). CDADNet contains three components: an attentional module, a context-guided module and a dense attentional dilated module. The attentional module is used to provide attention maps which can remove background information, while the context-guided module is proposed to extract multi-scale contextual information. Moreover, the dense attentional dilated module aims to generate high-granularity density maps and the cascaded strategy is used to preserve information from changing scales. To verify the feasibility of our method, we compare it to the existing approaches on five crowd counting datasets (ShanghaiTech (Part_A and Part_B), WorldEXPO’10, UCSD, UCF_CC_50). The comparison results demonstrate that CDADNet is effective and robust for various scenes.     

Performance evaluation of routing protocols for MANETs with known connectivity patterns using evolving graphs

The assessment of routing protocols for mobile wireless networks is a difficult task, because of the networks’ dynamic behavior and the absence of benchmarks. However, some of these networks, such as intermittent wireless sensors networks, periodic or cyclic networks, and some delay tolerant networks (DTNs), have more predictable dynamics, as the temporal variations in the network topology can be considered as deterministic, which may make them easier to study. Recently, a graph theoretic model—the evolving graphs—was proposed to help capture the dynamic behavior of such networks, in view of the construction of least cost routing and other algorithms. The algorithms and insights obtained through this model are theoretically very efficient and intriguing. However, there is no study about the use of such theoretical results into practical situations. Therefore, the objective of our work is to analyze the applicability of the evolving graph theory in the construction of efficient routing protocols in realistic scenarios. In this paper, we use the NS2 network simulator to first implement an evolving graph based routing protocol, and then to use it as a benchmark when comparing the four major ad hoc routing protocols (AODV, DSR, OLSR and DSDV). Interestingly, our experiments show that evolving graphs have the potential to be an effective and powerful tool in the development and analysis of algorithms for dynamic networks, with predictable dynamics at least. In order to make this model widely applicable, however, some practical issues still have to be addressed and incorporated into the model, like adaptive algorithms. We also discuss such issues in this paper, as a result of our experience.     

The expansion-gap problem in programs for trajectory analysis of dense electron beams

It is shown that the computational algorithms for calculating the region near the cathode and the expansion gap that are used in trajectory-analysis computer programs, do not match the accepted hydrodynamic-beam model or the conditions for emission in the ρ and T operation modes with zero initial velocity forming part of this model. As a consequence, the results of calculations of the beams with a high electron-beam compression of L ≥ 20 are invalid.     

Representing images using points on image surfaces.

This paper presents a new approach to represent an image by "verge points," which are defined as high-curvature points on the image surface. This representation offers a compact and reversible way to preserve the essence of the original image. Various applications, such as compression, edge detection, image enhancement, and image editing, can be achieved based on this representation. In this paper, the whole procedure for verge point representation is presented. Based on these verge points, image reconstruction can be easily achieved via iterative linear interpolation. These extracted verge points with compatible properties are further linked into verge curves to offer more compact representation. Progressive representation is also developed based on a multiscale extraction scheme. Some potential applications are then presented to demonstrate the versatility of this representation.     

基于C#与OpenGL的密集井井眼轨迹三维可视化系统研究

对海上钻井平台施工过程中的问题进行了调查,发现一个海洋平台上要求钻出越来越多的密集井,由于平台位置有限导致这些密集的井眼轨迹之间间距小,垂向可利用空间不足,加大了钻井工程设计、施工的风险,故对满足该环境下的井眼轨迹三维可视化系统进行研究.首先对井眼轨迹三维可视化系统中的多种井眼轨迹计算方法进行研究对比和分析,选择最小曲率法和曲率半径法作为主要的计算方法,根据计算的方法计算得到井眼轨迹数据,然后对井眼轨迹的三维建模方法展开分析对比.最后对井眼轨迹的三维可视化系统的结构进行总体设计,并选择C#语言和OpenGL技术作为开发的主要工具.     

Recognizing jump patterns with physics-based validation in human moving trajectory

This paper presents an approach to recognize jump patterns in human moving trajectory, differentiating jump tracks from planar moving tracks. Since human moving trajectory is one of the most informative representations for content understanding and event detection, trajectory-based video analysis has been gaining popularity. However, a jump action typically leads to violent change in human moving trajectory, since the person suddenly leaves the original plane on which he/she has been moving. The abnormal tracks of the trajectory would influence the performance of trajectory-based video analysis. Hence, differentiating jump tracks from planar moving tracks is of vital importance, not to mention that jump actions typically imply significant events, especially in sports games. In this paper, volleyball videos are used as case study to demonstrate the effectiveness of our proposed jump pattern recognition approach. We derive player trajectory by head tracking, analyze the movement of each player, and recognize potential jump tracks in player trajectories based on two important characteristics: (1) jumps cause pulse-like tracks in the trajectory and (2) the extensions of such tracks go through the vanishing point of vertical lines in the scenes. Finally, the jump positions/heights are estimated, in addition to the planar moving trajectory of each player on the court ground. The experiments show that satisfactory results can be obtained with the proposed recognition scheme.     

Proactive key distribution using neighbor graphs

User mobility in wireless data networks is increasing because of technological advances, and the desire for voice and multimedia applications. These applications, however, require that handoffs between base stations (or access points) be fast to maintain the quality of the connections. In this article we introduce a novel data structure, the neighbor graph, that dynamically captures the mobility topology of a wireless network. We show how neighbor graphs can be utilized to obtain a 99 percent reduction in the authentication time of an IEEE 802.11 handoff (full EAP-TLS) by proactively distributing necessary key material one hop ahead of the mobile user. We also present a reactive method for fast authentication that requires only firmware changes to access points and hence can easily be deployed on existing wireless networks.     

Mobile position location using factor graphs

Making use of the time-of-arrival measurements and their stochastic properties, we propose a low-complexity high-accuracy algorithm to estimate the location of a target mobile station (MS). Under a factor graph framework, in the proposed algorithm, soft-information is efficiently exchanged among local processing units to iteratively purify the estimate of the MS location. Numerical results show that the proposed algorithm not only enjoys advantages of low complexity, suitable for integrated-circuit implementation, but it is also able to achieve performance very close to the optimum achievable bound, the maximum-likelihood bound.     

Attributed skeleton graphs using mathematical morphology

The aim of the work described is to use the potential strength of the skeleton of discrete objects in computer vision and pattern recognition where features of objects are needed for classification. Algorithms are introduced for detecting skeleton characteristic points (end points, junction points and curve points) and creating an attributed graph, that can then be used as an input to graph matching algorithms, based on a morphological approach     

Lossy distributed source coding using graphs

In this letter, lossy distributed source coding using graphs is considered. This corresponds to the source coding part of the graph-based framework for transmission of analog correlated sources over the multiple-access channel (MAC). Consequently, it is shown that a pair of analog correlated sources can be reliably represented into a bipartite graph by allowing certain amount of distortion. An achievable rate-distortion region for this problem is also provided. Therefore, it can be concluded that, for transmission of any (both discrete and continuous) set of correlated sources over MACs, graphs can be used as discrete interface between source coding and channel coding modules.     

Radiation efficiency of dense antenna arrays forming flat-top and contour radiation patterns

Infinite 1D and 2D periodic planar arrays of densely spaced radiators that are used to form flat-top and contour patterns are analyzed. It is proposed to determine the amplitude distributions over the antenna aperture as the Fourier transforms of the window functions corresponding to the respective patterns. The proposed distributions are used to derive the explicit expressions showing that the efficiency of an array with a specified flat-top or contour pattern is equal to the mean efficiency of a uniformly excited array in the scan region covered by the specified pattern.     

Secure group communications using key graphs

Many emerging network applications are based upon a group communications model. As a result, securing group communications, i.e., providing confidentiality, authenticity, and integrity of messages delivered between group members, will become a critical networking issue. We present, in this paper, a novel solution to the scalability problem of group/multicast key management. We formalize the notion of a secure group as a triple (U,K,R) where U denotes a set of users, K a set of keys held by the users, and R a user-key relation. We then introduce key graphs to specify secure groups. For a special class of key graphs, we present three strategies for securely distributing rekey messages after a join/leave and specify protocols for joining and leaving a secure group. The rekeying strategies and join/leave protocols are implemented in a prototype key server we have built. We present measurement results from experiments and discuss performance comparisons. We show that our group key management service, using any of the three rekeying strategies, is scalable to large groups with frequent joins and leaves. In particular, the average measured processing time per join/leave increases linearly with the logarithm of group size     

Asymptotic trajectory tracking of manipulators using uncalibrated visual feedback

To implement a position-based visual feedback controller for a manipulator, it is necessary to calibrate the homogeneous transformation matrix between its base frame and the vision frame besides intrinsic parameters of the vision system. The accuracy of such a calibration greatly affects the control performance. Substantial efforts must be made to obtain a highly accurate transformation matrix. In this paper, we propose an adaptive visual feedback controller for manipulators when the homogeneous transformation matrix is not calibrated. It is assumed that the vision system can measure the 3D position and orientation of the manipulator in real-time. Based on an important observation that the unknown transformation matrix can be separated from the visual Jacobian matrix, we propose an adaptive algorithm, similar to the model-based adaptive algorithm, to estimate the unknown matrix online. The use of the proposed visual feedback controller greatly simplifies the implementation of a manipulator-vision workcell. This controller is especially useful when such a pre-calibration is not possible. It is proved by Lyapunov approach that the motion of the manipulator approaches asymptotically to the desired trajectory. Simulations and experimental results are included to demonstrate performance of this adaptive visual feedback controller.     

基于稠密轨迹特征的红外人体行为识别

提出了一种使用基于稠密轨迹(DT)融合特征的红外人体行为识别(HAR)方法。主要流程如下:1)通过稠密采样获得输入行为视频的DT;2)计算DT的方向梯度直方图(HOG)、光流直方图(HOF)和运动边界描述子(MBH)3个描述子;3)基于DT的HOG、HOF和MBH,并采取词袋库模型和一定的融合策略,构建融合特征;4)以第3步所构建的融合特征为k近邻分类器(k-NN)的输入,完成人体HAR。实验以IADB红外行为库为研究对象,正确识别率达到96.7%。结果表明,提出的特征融合及识别方法能有效地对红外人体行为进行识别。     

Improved low-density parity-check codes using irregular graphs

We construct new families of error-correcting codes based on Gallager's (1973) low-density parity-check codes. We improve on Gallager's results by introducing irregular parity-check matrices and a new rigorous analysis of hard-decision decoding of these codes. We also provide efficient methods for finding good irregular structures for such decoding algorithms. Our rigorous analysis based on martingales, our methodology for constructing good irregular codes, and the demonstration that irregular structure improves performance constitute key points of our contribution. We also consider irregular codes under belief propagation. We report the results of experiments testing the efficacy of irregular codes on both binary-symmetric and Gaussian channels. For example, using belief propagation, for rate 1/4 codes on 16000 bits over a binary-symmetric channel, previous low-density parity-check codes can correct up to approximately 16% errors, while our codes correct over 17%. In some cases our results come very close to reported results for turbo codes, suggesting that variations of irregular low density parity-check codes may be able to match or beat turbo code performance     

Saliency detection in human crowd images of different density levels using attention mechanism

The human visual system has the ability to rapidly identify and redirect attention to important visual information in high complexity scenes such as the human crowd. Saliency prediction in the human crowd scene is the process using computer vision techniques to imitate the human visual system, predicting which areas in a human crowd scene may attract human attention. However, it is a challenging task to identify which factors may attract human attention due to the high complexity of the human crowd scene. In this work, we propose Multiscale DenseNet — Dilated and Attention (MSDense-DAt), a convolutional neural network (CNN) using self-attention to integrate the result of knowledge-driven gaze in the human visual system to identify salient areas in the human crowd scene. Our method combines various state-of-the-art deep learning architectures to deal with the high complexity in human crowd image, such as multiscale DenseNet for multiscale deep features extraction, self-attention, and dilated convolution. Then the effectiveness of each component in our CNN architecture is evaluated by comparing different components combinations. Finally, the proposed method is further evaluated in different crowd density levels to appraise the effect of crowd density on model performance.     

Unified design of iterative receivers using factor graphs

Iterative algorithms are an attractive approach to approximating optimal, but high-complexity, joint channel estimation and decoding receivers for communication systems. We present a unified approach based on factor graphs for deriving iterative message-passing receiver algorithms for channel estimation and decoding. For many common channels, it is easy to find simple graphical models that lead directly to implementable algorithms. Canonical distributions provide a new, general framework for handling continuous variables. Example receiver designs for Rayleigh fading channels with block or Markov memory, and multipath fading channels with fixed unknown coefficients illustrate the effectiveness of our approach