Great progress has been made in the integration of Unmanned Aerial Vehicle (UAV) magnetic measurement systems, but the interpretation of UAV magnetic data is facing serious challenges. This paper presents a complete workflow for the detection of the subsurface objects, like Unexploded Ordnance (UXO), by the UAV-borne magnetic survey. The elimination of interference field generated by the drone and an improved Euler deconvolution are emphasized. The quality of UAV magnetic data is limited by the UAV interference field. A compensation method based on the signal correlation is proposed to remove the UAV interference field, which lays the foundation for the subsequent interpretation of UAV magnetic data. An improved Euler deconvolution is developed to estimate the location of underground targets automatically, which is the combination of YOLOv3 (You Only Look Once version 3) and Euler deconvolution. YOLOv3 is a deep convolutional neural network (DCNN)-based image and video detector and it is applied in the context of magnetic survey for the first time, replacing the traditional sliding window. The improved algorithm is more satisfactory for the large-scale UAV-borne magnetic survey because of the simpler and faster workflow, compared with the traditional sliding window (SW)-based Euler method. The field test is conducted and the experimental results show that all procedures in the designed routine is reasonable and effective. The UAV interference field is suppressed significantly with root mean square error 0.5391 nT and the improved Euler deconvolution outperforms the SW Euler deconvolution in terms of positioning accuracy and reducing false targets.

For unexploded O=ordnance (UXO) detection, individual technology cannot achieve the best detection performance. The new detection mode of joint magnetic and electromagnetic method has attracted more and more attention. In this paper, a newly developed joint detection system is introduced, a multi-rotor UAV-based magnetic system (UAVMAG) and a cart-based time-domain electromagnetic detection system (TDEM-Cart) are combined, and the cooperative processing of magnetic field and electromagnetic data is proposed. The result of the joint inversion fuses the feature vector retrieved from the magnetic field data and the feature vector inverted from the electromagnetic data, providing more accurate positioning results and richer information, which is favorable to locate and distinguish the UXO. Two field experiments are conducted, and the results show that when the joint system works in the full-coverage survey mode, both ferromagnetic and non-ferromagnetic metal targets can be detected, avoiding missed detections. In addition, when the joint system works in the cued survey mode, the detection efficiency is improved, the positioning accuracy of joint interpretation is less than 10 cm, and it shows satisfactory performance in the recognition of targets.

With the development of deep neural networks, multi-channel speech separation techniques with fixed array geometries have achieved remarkable performance. However, distributed microphone array processing remains a challenging problem because it requires the network to be able to process inputs with varying dimensions. To address this problem, we propose a triple-path recurrent neural network (TPRNN) with multi-scale aggregation blocks for distributed microphone array multi-channel speech separation. First, we extend the single-channel dual-path recurrent neural network by additionally adding multi-scale aggregation blocks and adaptive feature fusion blocks. Next, a third path along the spatial dimension is introduced to model spatial information. By this means, TPRNN can alternately and iteratively perform inter-channel, intra-chunk, and inter-chunk modeling. The experimental results show that the proposed approach outperforms other advanced baselines for multi-channel speech separation and enhancement tasks using spatially distributed microphones.

The interference of sidelobe often causes different targets to exhibit similar features, diminishing fine-grained classification accuracy. This effect is particularly pronounced when the available data are limited. To address the aforementioned issues, a novel classification framework for sidelobe-affected SAR imagery is proposed. First, a method based on maximum median filtering is adopted to remove sidelobe by exploiting local grayscale differences between the target and sidelobe, constructing a high-quality SAR dataset. Second, a deep metric learning network is constructed for fine-grained classification. To enhance the classification performance of the network on limited samples, a feature extraction module integrating a lightweight attention mechanism is designed to extract discriminative features. Then, a hybrid loss function is proposed to strengthen intra-class correlation and inter-class separability. Experimental results based on the FUSAR-Ship dataset demonstrate that the method exhibits excellent sidelobe suppression performance. Furthermore, the proposed framework achieves an accuracy of 84.18% across five ship target classification categories, outperforming the existing methods, significantly enhancing the classification performance in the context of sidelobe interference and limited datasets.

Thanks to the use of deep neural networks (DNNs), microphone array speech separation methods have achieved impressive performance. However, most existing neural beamforming methods explicitly follow traditional beamformer formulas, which possibly causes sub-optimal performance. In this study, a pre-separation and all-neural beamformer framework is proposed for multi-channel speech separation without following the solutions of the conventional beamformers, such as the minimum variance distortionless response (MVDR) beamformer. More specifically, the proposed framework includes two modules, namely the pre-separation module and the all-neural beamforming module. The pre-separation module is used to obtain pre-separated speech and interference, which are further utilized by the all-neural beamforming module to obtain frame-level beamforming weights without computing the spatial covariance matrices. The evaluation results of the multi-channel speech separation tasks, including speech enhancement subtasks and speaker separation subtasks, demonstrate that the proposed method is more effective than several advanced baselines. Furthermore, this method can be used for symmetrical stereo speech.