In this paper, we construct a lightweight, high-precision and high-speed object tracking using a trained CNN. Conventional methods with trained CNNs use VGG16 network which requires powerful computational resources. Therefore, there is a problem that it is difficult to apply in low computation resources environments. To solve this problem, we use MobileNetV3, which is a CNN for mobile terminals.Based on Feature Map Selection Tracking, we propose a new architecture that extracts effective features of MobileNet for object tracking. The architecture requires no online learning but only offline learning. In addition, by using features of objects other than tracking target, the features of tracking target are extracted more efficiently. We measure the tracking accuracy with Visual Tracker Benchmark and confirm that the proposed method can perform high-precision and high-speed calculation even in low computation resource environments.
翻译:在本文中,我们使用训练有素的CNN, 构建了一个轻量级、高精度和高速的物体跟踪。有受过训练的CNN的常规方法使用VGG16网络,这需要强大的计算资源。因此,在低计算资源环境中很难应用这个问题。为了解决这个问题,我们使用移动网络3,这是移动终端的CNN。基于特征地图选择跟踪,我们提出了一个新的架构,为对象跟踪提取移动网络的有效功能。这一架构不需要在线学习,而只需要离线学习。此外,通过使用跟踪目标以外的对象的特征,跟踪目标的特征被更高效地提取。我们用视觉追踪基准测量跟踪准确性,并确认拟议方法即使在低计算资源环境下也能进行高精度和高速计算。