Various observation models have been introduced into the object tracking community, and combining them has become a promising direction. This paper proposes a novel approach for estimating the confidences of different observation models, and then effectively combining them in the particle filter framework. In our approach, spatial Likelihood distribution is represented by three simple but efficient parameters, reflecting the overall similarity, distribution sharpness and degree of multi peak. The balance of these three aspects leads to good estimation of confidences, which helps maintain the advantages of each observation model and further increases robustness to partial occlusion. Experiments on challenging video sequences demonstrate the effectiveness of our approach.

Object detection and tracking is one of the most important research topics in pattern recognition and the basis of many computer vision systems. Many accomplishments in this field have been achieved recently. Some specific objects, such as human face and vehicles, can already be detected in various applications. However, tracking objects with large variances in color, texture and local shape (such as pedestrians) is still a challenging topic in this field. To solve this problem, a pedestrian tracking scheme is proposed in this paper, including online training for pedestrian-detector. Simulation and analysis of the results shows that, the proposal method could deal with illumination change, pose change and occlusion problem and any combination thereof.

Heterogeneous networks (HetNets) are emerging as an inevitable method to tackle the capacity crunch of the cellular networks. Due to the complicated network environment and a large number of configured parameters, coverage and capacity optimization (CCO) is a challenging issue in heterogeneous cellular networks. By combining the self-optimizing algorithm for radio frequency (RF) parameters with the power control mechanism of small cells, the CCO problem of self-organizing network is addressed in this paper. First, the optimization of RF parameters is solved based on reinforcement learning (RL), where the base station is modeled as an agent that can learn effective strategies to control the tunable parameters by interacting with the surrounding environment. Second, the small cell can autonomously change the state of wireless transmission by comparing its distance from the user equipment with the virtual cell size. Simulation results show that the proposed algorithm can achieve better performance on user throughput compared to different conventional methods.

Smartphones for wireless communication typically consist of a large area frontal liquid crystal display (LCD), which incorporates a metal back plate, and a back cover chassis made from metal. Leveraging this structure a new approach to construct antennas for smartphones is proposed where the complete metal back cover chassis and LCD back plate are used as the radiating element and ground plane. In the design a feedline is connected between the metal back cover chassis and LCD back plate, along with shorts at various locations between the two metal plates, to control the resonance frequency of the resulting antenna. Multiple-band operation is possible without the need for any slots in the plates for radiation. Results show that antenna frequency reconfigurability can be achieved when switching function is added to the shorts so that several wireless communication bands can be covered. This approach is different from existing metallic frame antenna designs currently available in the market. A design example is provided which uses one PIN diode for the switching shorts and the target frequency bands are 740-780MHz and 900-1000MHz & 1700-1900MHz. The optimization of LC matchings and concerns of hand effects and metallic components between the chassis and LCD metal back plate are also addressed.