27 October 2023 in IEEE Transactions on Instrumentation and Measurement
Drone audio detection methods have become a key component of anti-drone systems. Traditional audio feature extraction methods have problems such as large fluctuations in feature vectors, fixed extraction resolution, and redundant feature information extraction. Moreover, the dataset is ideal and not representative of real application scenarios. In this study, a dual-domain audio feature extraction method in the time and frequency domains is proposed that improves the accuracy of drone detection by combining the more richly detailed information in the time domain and the relatively stable property of the signal in the frequency domain. A real-world sound dataset that contains low signal-to-noise ratio audio was collected for experimental validation. The results showed that, compared with existing methods, the proposed method took full advantage of the “zoom” feature of the wavelet packet transform, the local feature extraction capability of a one-dimensional convolutional neural network, and the global modeling capability of a self-attention mechanism, thereby effectively improving the success rate of drone detection in common scenarios. The proposed method also outperformed other methods with respect to several evaluation metrics.
8 February 2024 in Physica Scripta
In a tactical warfare setting, the efficacy of target detection becomes profoundly compromised due to prevalent environmental factors such as smoke, dust, and atmospheric interference. Such impediments invariably undermine the precision and reliability of identifying pivotal targets, thereby precipitating potentially dire ramifications. Remarkably, short-wave infrared technology has exhibited unparalleled proficiency in elucidating target attributes even amidst challenging conditions characterized by smoke, fog, or haze. Against this backdrop, the present study delineates a pioneering algorithmic framework that seamlessly amalgamates the imperatives of image registration and fusion. This is achieved through the deployment of an advanced dual-discriminator Generative Adversarial Network (GAN), specifically tailored for amalgamating short-wave infrared and visible light imagery within smoke-obscured contexts. Our methodology commences with the introduction of an augmented Speeded-Up Robust Features (SURF) algorithm, meticulously designed to rectify inherent misalignments within the input imagery. Subsequent enhancements encompass the refinement of the generator’s loss function and the integration of a multi-scale convolutional kernel, thereby facilitating the extraction and amalgamation of a more expansive array of salient features. This concerted effort culminates in the elevation of image fusion quality. To corroborate the efficacy and robustness of our proposed framework, rigorous validation procedures were conducted utilizing a meticulously curated dataset comprising short-wave infrared and visible light images. Empirical evaluations, encompassing both subjective and objective comparative analyses, unequivocally affirm the superior performance metrics of our fusion network. Specifically, our methodology surpasses alternative fusion techniques across multiple dimensions, including visual fidelity, perceptual quality, and structural congruence of synthesized images.