JDE multi-object tracking algorithm integrating multi-level semantic enhancement

Abstract

In order to solve the problem of target ID switching caused by target occlusion and insufficient ID information and location information extraction in JDE(joint detection and embedding) algorithm, an improved multi-target tracking algorithm based on JDE is proposed in this paper. Firstly, the SPA feature space pyramid attention module is used to expand the receptive field and obtain more abundant semantic information to improve the detection accuracy of the model for different scale targets. Secondly, the FCN network makes the header and ID Embedding task collaborative learning to alleviate the excessive competition and enhance the original semantic information, effectively reducing the number of ID switching. Finally, PCCs-Ma motion measurement can strengthen the connection between Kalman filtering prediction and observation, and improve the reliability of similarity discrimination of motion characteristics. In order to verify the effectiveness of the algorithm, the JDE algorithm and the proposed algorithm are compared in the same experimental environment. The experimental results show that the average accuracy of model detection is improved by 3.94 %. On the MOT16 dataset, the MOTA and IDF1 indexes are increased by 6.9 %, and the number of ID switching of the improved algorithm is significantly reduced, and good tracking results are achieved.

A Survey of UAV Swarm Task Allocation Based on the Perspective of Coalition Formation

“…The UAV swarm first determines the task priority according to the relevant requirements of the task execution and then selects the global optimal UAV for the task with the highest current priority in the task priority sequence in turn. As shown in Figure 4, when the UAV carries insufficient resources to complete the strike task of the target alone, the UAV becomes the initiator of coalition formation and broadcasts the specific information of the discovered target to other UAVs in the swarm (Wang et al, 2005). Then, other UAVs respond to the coalition initiator according to their own resources.…”

“…In order to avoid coalition deadlock and resource waste, Wang et al (2005) proposed a coalition formation strategy based on an improved task matching calculation method. When dealing with the current task, this method can selectively learn from the historical experience of the system and give full play to its learning ability.…”

Optimization of the Air Distribution in a Biomass Grate-Fired Furnace

“…Ji et al used natural gas to simulate the volatiles released from coal and investigated the unsteady combustion mechanism of an industrial moving grate boiler under two typical air distribution modes [15]. Zhao et al conducted a numerical simulation study using FLIC + Fluent [16] software to analyze the combustion characteristics of straw direct-firing boilers at a specific load, only changing the ratio of primary and secondary air [17]. However, due to the difference in combustion characteristics between coal and biomass, there is limited research on air distribution modes in grate-firing boilers for biomass combustion.…”