The increasing use of UAVs highlights the need for improved detection methods to manage low-altitude airspace. However, existing detection techniques often struggle to address the challenge of identifying small fast-moving objects in complex backgrounds. To address these issues, we develop an innovative ensemble detection method called DAM (Dual Attention Mechanism), which enhances the YOLOv8 (You Look Only Once) algorithm by integrating Swin Transformer and Global Attention Mechanism (GAM). Our proposed method is a comprehensive solution that includes several key advancements. Firstly, DAM incorporates the Swin Transformer to capture a wider context for more accurate object recognition. Secondly, the global attention mechanism was used to focus on the salient features in the detection framework. Thirdly, the dataset BFAIRD (Battlefield Air Drones) from the game Battlefield V is introduced to provide a realistic test for the model. The experimental results show that the integration of these components into the DAM model is better than the performance indicators such as precision, recall and mean average precision (mAP) in various scenarios compared with YOLOv8. The code of our model is available on Github for further exploration and application.