MobileTrack: Siamese efficient mobile network for high‐speed UAV tracking

Recently, Siamese-based trackers have drawn amounts of attention in visual tracking field because their excellent performance. However, object on Unmanned Aerial Vehicles platform encounters difficulties under circumstances such as small objects and similar interference. Most existing methods for aerial adopt deeper networks or inefficient policies to promote performance, but most can hardly meet real-time requirements mobile platforms with limited computing resources. Thus, this work, an efficient lightweight siamese tracker (MobileTrack) is proposed high-time tracking, realising the balance between performance speed. Firstly, a convolutional network (D-MobileNet) designed enhance characterisation ability objects. Secondly, object-aware module local cross-channel information exchange, enhancing feature object. Besides, anchor-free region proposal introduced predict pixel by pixel. Finally, deep shallow fully utilised cascading multiple accurate locating robust tracking. Extensive experiments three benchmarks show that achieves outstanding while keeping beyond-real-time