Qiang Wang

Although recent multi-object tracking (MOT) methods have achieved considerable progress, tracking from a moving vehicle remains challenging because ego-motion, scale variation, occlusion, and uncertain detections degrade motion prediction and data association. To address these issues, we propose an adaptive motion association framework consisting of a task-aligned enhancement head (TAE-Head), adaptive motion estimation (AME), and extended shape-constrained IoU (ES-IoU). TAE-Head improves the consistency between classification and localization features. AME adjusts process and measurement covariances, estimates a confidence-weighted common displacement from provisional matches, corrects predicted positions, and recomputes association before the final state update. ES-IoU increases geometric tolerance while suppressing shape-incompatible matches. On BDD100K, the proposed framework achieves 56.7% HOTA and 65.0% IDF1, exceeding the baseline by 9.0 and 7.7 percentage points, respectively, while operating at 32.5 FPS. On MOT17, it achieves 68.2% HOTA and 80.5% IDF1.

Timely collecting logistics information and finding anomalies of material supply plays a critical role in modern manufacturing systems. The problem is how to obtain multi-attribute logistics information of production logistics and build an effective approach for mining anomalies from the huge number of RFID data. The multi-attribute, randomness and various measure units of logistics states further aggravate the problem. In this paper, a novel RFID-based logistics information processing approach is proposed. Firstly, the state features of production logistics is discussed from multi-attribute perspectives including time, location, quantities, sequence and path, and a set of calculating models is set up to process RFID data for getting multi-attribute state data. Furthermore, in case of the randomness and various measure units of state data, a similarity model is presented to unify measure units of state data, and a clustering approach is proposed to divide the huge number of RFID data into different clusters with high close degree for finding out anomalies. Lastly, the experimental results show that the proposed approach can efficiently find out more than 90% of anomalies among production logistics.

This paper studies the fault-tolerant cooperative control problem of agent groups in the context of multi-agent flocking tasks with second order linear dynamics. A distributed flocking algorithm with biconnected topology is proposed which is composed of two parts: motion strategy of biconnectivity and fault-tolerant flocking algorithm with bounded control input. The proposed algorithm handles the movement and reconfiguration of the flock, while maintaining the desired shape. It is proved that the proposed control algorithm can not only achieve the resultant biconnected network which is able to tolerate temporary node failures, but also guarantee the stable flocking motion. Several simulations are presented to demonstrate the efficiency of the theoretical results.