Despite extensive research dedicated to improving the accuracy and efficiency of deep learning-based UAV intent recognition, its practical application in real-world UAV situation recognition scenarios is hindered by limited accuracy and lack of realism. In this article, we propose a novel generative model for intent recognition that leverages the standard transformer architecture along with pretraining and initialization techniques to achieve superior accuracy and practicality in UAV intent recognition. Experimental results on databases with uneven distributions demonstrate that our proposed model achieves an accuracy of approximately 79%, surpassing traditional classical intent classification networks by 13%–15%. Further experimentation on the network security laboratory-knowledge discovery in databases (NSL-KDD) database corroborates the broad applicability of the proposed generative method in addressing database distribution imbalances. Moreover, our proposed UAV intent recognition ...

This article introduces a novel cross-modal neural network model that aims to convert long-term temporal behavior data into natural language to achieve unmanned aerial vehicle (UAV) intent recognition. Our generative intent recognition model effectively utilizes the inherent redundancy present in long temporal behavior data by incorporating a sequence compression module, which enables the cross-modal generation and alignment of intents while preserving the integrity of the standard Transformer architecture. Importantly, we observe that this approach mitigates the negative impact of imbalanced database distribution by mapping intent categories onto the modality of natural language. Furthermore, we propose three comprehensive pretraining tasks specifically designed for time series data, thoroughly examining their interconnections and analyzing the impact of a hybrid pretraining framework on the accuracy of intent recognition. Our experimental results demonstrate the superiority of our pr...