International Journal of Scientific Research and Engineering Development

In dynamic scenes, object tracking faces challenges such as complex background interference and rapid object movement. To explore the application of object tracking technology in dynamic scenes, this paper utilises a visual model based on LSTM-Attention, aiming to improve the accuracy of object tracking in dynamic scenes. This model combines the memory capabilities of LSTM in processing time-series data with the advantages of the Attention mechanism in capturing the dynamic characteristics of objects in dynamic scenes. Building on this foundation, the model performs object tracking tasks in dynamic scenes. Its performance is compared with that of traditional object tracking algorithms, and the impact of various scene factors on tracking results is analysed to validate the advantages of the new model. Experimental data indicate that the model can accurately and stably track objects even when they are moving rapidly or partially occluded, maintaining an accuracy of 80.4% in complex dynamic scenes and demonstrating strong resistance to interference.

Sun Chenyang,"A Study on Object Tracking in Dynamic Scenes Using an LSTM-Attention Model" International Journal of Scientific Research and Engineering Development, V9(2): Page(1832-1835) Mar-Apr 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.