Accurate P-wave first-motion-polarity (FMP) information can contribute to solving earthquake focal mechanisms, especially for small earthquakes, to which waveform-based methods are generally inapplicable due to the computationally expensive high-frequency waveform simulations and inaccurate velocity models. In this paper, we propose a deep-learning-based method for the automatic determination of the FMPs, named “DiTingMotion”. DiTingMotion was trained with the P-wave FMP labels from the “DiTing” and SCSN-FMP datasets, and it achieved ∼97.8% accuracy on both datasets. The model maintains ∼83% accuracy on data labeled as “Emergent”, of which the FMP labels are challenging to identify for seismic analysts. Integrated with HASH, we developed a workflow for automated focal mechanism inversion using the FMPs identified by DiTingMotion and applied it to the 2019 M 6.4 Ridgecrest earthquake sequence for performance evaluation. In this case, DiTingMotion yields comparable focal mechanism results to that using manually determined FMPs by SCSN on the same data. The results proved that the DiTingMotion has a good generalization ability and broad application prospect in rapid earthquake focal mechanism inversion.
An approximate 236‐year interval in major seismic activity near the southern Xianshuihe fault terminated on 5 September 2022 until a destructive M6.8 Luding earthquake. The strike‐slip mainshock, accompanied by potent normal‐faulting aftershocks, fell short of the previously anticipated M7+ event. Utilizing deep‐learning and source inversion techniques, we analyzed dense near‐field seismograms to examine detailed fault structures and rupture characteristics. Our high‐quality relocated catalog with 7,388 events revealed that the earthquake ruptured an unmapped multiscale fault network. The analysis of the mainshock and 43 ML ≥ 2.8 aftershocks suggests normal‐faulting events relate to the vertical movement of the Gongga Mountain. The sequence length, seismic gaps, and asperities derived from seismicity and waveform modeling imply that the rupture of the M6.8 quake was incomplete, indicating a high risk of a future M6+ event at the northern Moxi segment. These findings hold crucial implications for assessing future seismic hazards in this region.
20 January 2022 in IEEE Geoscience and Remote Sensing Letters
Recently, deep learning methods have been successfully applied to the ship detection in synthetic aperture radar (SAR) images. It is still a great challenge to detect SAR ships, due to the extremely poor image quality and complex background. To solve the problems, a novel method named orientation-aware feature fusion network (OFF-Net) for ship detection in SAR images is proposed in this letter. OFF-Net consists of global context path aggregation (GCPA) module and local rotated contrast enhance (LRCE) module, which fuses the global and local information in feature extraction. First, GCPA module is explored to integrate the global context block with path aggregation network (PAN) to learn the global background information. Second, by designing a rotation scheme based on feature map cyclic shift with four directions, LRCE module is developed to enhance the targets and suppress the background clutters in SAR images. Finally, a decoupled orientation-aware head is proposed to handle the arbitrarily rotated ships more robustly and alleviate the conflict between classification and regression tasks during detection. In addition, a high-resolution SAR-ship detection dataset (OBB-HRSDD) with rotatable bounding boxes is provided. The detection results on the SAR ship detection dataset (SSDD+) and OBB-HRSDD illustrate that our method outperforms all the compared methods. The code and OBB-HRSDD are released at https://github.com/SJX152/papercode
1 February 2022 in International Journal of Remote Sensing
The implementation of multi-modal remote sensing image registration is still a challenge in various applications. In this paper, a novel framework called Heterogeneous SuperPoint Network is proposed for multi-modal remote sensing image registration. Heterogeneous SuperPoint Network operates on full-sized images and produces multi-modal interest points accompanied with fixed length descriptors in a single forward pass. Moreover, we develop a strategy for training a base detector that used in conjunction with heterogeneous homographic adaptation to generate pseudo-ground truth interest point labels for unlabeled images in a self-supervised fashion. Then, the interest points and point correspondences obtained by Heterogeneous SuperPoint Network are used to do the final matching with the nearest neighbor (NN) matching and Random sample consensus (RANSAC). Various image sets have been considered in this paper for the evaluation of the proposed approach. Experimental results demonstrate the superior performance of the proposed method over the existing design for the whole data set.
The number of video clips available online is growing at a tremendous pace. Conventionally, user-supplied metadata text, such as the title of the video and a set of keywords, has been the only source of indexing information for user-uploaded videos. Automated extraction of video content for unconstrained and large scale video databases is a challenging and yet unsolved problem. In this paper, we present an audiovisual celebrity recognition system towards automatic tagging of unconstrained Web videos. Prior work on audiovisual person recognition relied on the fact that the person in the video is speaking and the features extracted from audio and visual domain are associated with each other throughout the video. However, this assumption is not valid on unconstrained Web videos. Proposed method finds the audiovisual mapping and hence improve upon the association assumption. Considering the scale of the application, all pieces of the system are trained automatically without any human supervision. We present the results on 26,000 videos and show the effectiveness of the method per-celebrity basis.