Object tracking algorithm based on improved siamese network
Computer vision technology is an interdisciplinary field involving computer science and image processing,and it is an important part of deep learning and artificial intelligence research.It also has great potential for engineering applications.Visual object tracking is an important research direction in the field of computer vision,widely used in areas such as autonomous driving,robots,and video surveillance.This paper proposes an improved Siam-ours object tracking algorithm for limited computing resources in unmanned platforms,aiming to reduce the number of algorithm parameters and computational complexity.The algo-rithm uses MobileNetV3 as the main network for feature extraction,and introduces the Focal Loss function to address the problem of imbalanced sample distribution in the training process caused by the original Contrastive Loss function,which leads to deteriora-tion of algorithm performance.The improved algorithm improves the accuracy and success rate by 1.7 and 2.3 percentage points,respectively,on the UAV123 test set.
computer visionunmanned platformMobileNetV3Focal Loss
万方数据
