Estimating directions of arrival (DOA) without knowledge of the source number is regarded as a challenging task, particularly when coherence among sources exists. Researchers have trained deep learning (DL)-based models to attack the problem of DOA estimation. However, existing DL-based methods for coherent sources do not adapt to variable source numbers or require signal independence. Herein, we put forward a new framework combining parallel DOA estimators with Toeplitz matrix reconstruction to address the problem. Each estimator is constructed by connecting a multi-label classifier to a spatial filter, which is based on convolutional-recurrent neural networks. Spatial filters divide the angle domain into several sectors, so that the following classifiers can extract the arrival directions. Assisted with Toeplitz-based method for source-number determination, pseudo or missed angles classified by the estimators will be reduced. Then, the spatial spectrum can be more accurately recovered. In addition, the proposed method is data-driven, so it is naturally immune to signal coherence. Simulation results demonstrate the predominance of the proposed method and show that the trained model is robust to imperfect circumstances such as limited snapshots, colored Gaussian noise, and array imperfections.
24 August 2016 in Rapid Communications in Mass Spectrometry
This letter puts forward a new method to estimate the direction of arrivals (DOAs) along with the source number under wideband circumstances. The method constructs a multiple-input–single-output (MISO) framework to make full use of the data from different frequency subbands. Each channel in the framework extracts angular features in parallel through stacked residual blocks. As a result, we can obtain an integrated output vector with smaller error. Because of the well-designed loss function, outputs of the trained model under the proposed framework are self-adjustable with the source number so that the regression of angle values and classification of signal numbers can be completed simultaneously. Moreover, we have proved herein that, if the equivalent matrices of residual blocks are negative definite, then the proposed framework is stable in the sense of Lyapunov. This stability is of great significance to practical applications. Numerical simulations on X-band data demonstrate high estimation accuracy and powerful generalization ability of the proposed scheme in the presence of nonuniform noise.
Road network is an important part of modern transportation. For the demands of accurate road information in practical applications such as urban planning and disaster assessment, we propose a multiscale method to extract road network from high-resolution synthetic aperture radar (SAR) images, which consists of three stages: potential road area segmentation, preliminary network generation, and road network refinement. Multiscale analysis is implemented using an image pyramid framework together with a fixed-size filter. First, a directional road detector is designed to highlight road targets in feature response maps. Subsequently, adaptive fusion is performed independently at each image scale, followed by a threshold method to produce potential road maps. Then, binary maps are decomposed according to the obtained direction information. For each connected component (CC), quality evaluation is conducted to further distinguish road segments and polynomial curve fitting is adopted as a thinning method. Multiscale information fusion is realized through the weighted sum of road curves. Finally, tensor voting and spatial regularization are employed to generate the final road network. Experiments on three TerraSAR images demonstrate the effectiveness of the proposed algorithm to extract road network completely and correctly.
Most direction‐of‐arrival (DOA) estimation models based on deep‐learning (DL) assume a fixed modulation type for array received signals. Once exposed to unknown signal types, the performance of these data‐driven models tends to degrade, posing a fundamental challenge that remains to be addressed. This paper puts forward a new rotation‐invariant (RI) DOA estimation framework based on deep residual learning (DRL), which shows strong adaptability across different signal modulation schemes. Moreover, the correlation between network weights and the underlying angular resolution is analyzed, serving as a reference for network scale design. Numerical simulations demonstrate the superiority of the proposed method over both traditional and DL‐based DOA estimation algorithms in the presence of impulsive noise.
Neurologists typically identify epileptic seizures from electroencephalograms (EEGs) by visual inspection. This process is often time-consuming, especially for EEG recordings that last hours or days. To expedite the process, a reliable, automated, and patient-independent seizure detector is essential. However, developing a patient-independent seizure detector is challenging as seizures exhibit diverse characteristics across patients and recording devices. In this study, we propose a patient-independent seizure detector to automatically detect seizures in both scalp EEG and intracranial EEG (iEEG). First, we deploy a convolutional neural network with transformers and belief matching loss to detect seizures in single-channel EEG segments. Next, we extract regional features from the channel-level outputs to detect seizures in multi-channel EEG segments. At last, we apply postprocessing filters to the segment-level outputs to determine seizures' start and end points in multi-channel EEGs. Finally, we introduce the minimum overlap evaluation scoring as an evaluation metric that accounts for minimum overlap between the detection and seizure, improving upon existing assessment metrics. We trained the seizure detector on the Temple University Hospital Seizure (TUH-SZ) dataset and evaluated it on five independent EEG datasets. We evaluate the systems with the following metrics: sensitivity (SEN), precision (PRE), and average and median false positive rate per hour (aFPR/h and mFPR/h). Across four adult scalp EEG and iEEG datasets, we obtained SEN of 0.617-1.00, PRE of 0.534-1.00, aFPR/h of 0.425-2.002, and mFPR/h of 0-1.003. The proposed seizure detector can detect seizures in adult EEGs and takes less than 15s for a 30 minutes EEG. Hence, this system could aid clinicians in reliably identifying seizures expeditiously, allocating more time for devising proper treatment.