Showing papers by "Shaobo Li published in 2022"

A Hybrid Matching Network for Fault Diagnosis under Different Working Conditions with Limited Data

Abstract: Intelligent fault diagnosis methods based on deep learning have achieved much progress in recent years. However, there are two major factors causing serious degradation of the performance of these algorithms in real industrial applications, i.e., limited labeled training data and complex working conditions. To solve these problems, this study proposed a domain generalization-based hybrid matching network utilizing a matching network to diagnose the faults using features encoded by an autoencoder. The main idea was to regularize the feature extractor of the network with an autoencoder in order to reduce the risk of overfitting with limited training samples. In addition, a training strategy using dropout with random changing rates on inputs was implemented to enhance the model's generalization on unseen domains. The proposed method was validated on two different datasets containing artificial and real faults. The results showed that considerable performance was achieved by the proposed method under cross-domain tasks with limited training samples.

A Real-Time Negative Obstacle Detection Method for Autonomous Trucks in Open-Pit Mines

Abstract: Negative obstacles such as potholes and road collapses on unstructured roads in open-pit mining areas seriously affect the safe transportation of autonomous trucks. In this paper, we propose a real-time negative obstacle detection method for self-driving trucks in open-pit mines. By analyzing the characteristics of road negative obstacles in open-pit mines, a real-time target detection model based on the Yolov4 network was built. It uses RepVGG as the backbone feature extraction network, applying SimAM space and a channel attention mechanism to negative obstacle multiscale feature fusion. In addition, the classification and prediction modules of the network are optimized to improve the accuracy with which it detects negative obstacle targets. A non-maximum suppression optimization algorithm (CIoU Soft Non-Maximum Suppression, CS-NMS) is proposed in the post-processing stage of negative obstacle detection. The CS-NMS calculates the confidence of each detection frame with weighted optimization to solve the problems of encountering obscure negative obstacles or poor positioning accuracy of the detection boxes. The experimental results show that this research method achieves 96.35% mAP for detecting negative obstacles on mining roads with a real-time detection speed of 69.3 fps, and that it can effectively identify negative obstacles on unstructured roads in open-pit mines with complex backgrounds.

Surface Extraction and Segmentation From 3-D Underwater Sub-Bottom Point Clouds Using Enhancement Filtering and Global Energy Optimization

Abstract: Nowadays, a 3-D sub-bottom profiler (SBP) can produce the point clouds of the subseabed and is gradually receiving more attention in getting geologically significant surfaces to reveal sedimentary environments and structural features. However, little literature studied the automatic extraction of these surfaces from the 3-D SBP data currently. Thus, this article proposes a hybrid method consisting of a surface extraction algorithm and a segmentation algorithm. First, the multiprofile SBP data are converted into 3-D data volume. Second, by taking full advantage of the plate-like characteristic of the layer surface in the 3-D SBP data, a plate-like enhancement filtering algorithm based on the nonuniform Gaussian scale is given to filter the 3-D data volume. Third, a threshold extraction is applied to extract surface voxels, and a hybrid region growing algorithm is put forward to segment surface voxels into basic units by combining multicriteria. Finally, the surface segmentation problem is formulated as global energy optimization, and a stepwise segmentation algorithm is proposed to get the final surface set. To verify the effectiveness of the proposed method, experiments were conducted and analyzed. The results showed that the proposed method performed well.

Seabed Sediment Classification Using Spatial Statistical Characteristics

Abstract: Conventional sediment classification methods based on Multibeam Echo System (MBES) data have low accuracy since the correlation between features and sediment has not been fully considered. Moreover, their poor resistance to the residual error of MBES backscatter strength (BS) processing also degrades their performances. Toward these problems, we propose a seabed sediment classification method using spatial statistical features extracted from angular response curve (ARC), topography, and geomorphology. First, to reduce interference of noise and residual error of beam pattern correction, we propose a robust method combining the Generic Seafloor Acoustic Backscatter (GSAB) model and Huber loss function to estimate the parameters of ARC which is strongly correlated with seabed sediments. Second, a feature set is constructed by AR features composed of GSAB parameters, BS mosaic and its derivatives, and seabed topography and its derivatives to characterize seabed sediments. After that, feature selection and probability map acquisition are employed based on the random forest algorithm (RF). Finally, a denoising and final sediment map generation method is proposed and applied to probability maps to obtain the sediment map with reasonable sediment distribution and clear boundaries between classes. We implement experiments and achieve the classification accuracy of 93.3%, which verifies the validity of our method.

Knowledge distillation for multi-depth-model-fusion recommendation algorithm

Abstract: Recommendation algorithms save a lot of valuable time for people to get the information they are interested in. However, the feature calculation and extraction process of each machine learning or deep learning recommendation algorithm are different, so how to obtain various features with different dimensions, i.e., how to integrate the advantages of each model and improve the model inference efficiency, becomes the focus of this paper. In this paper, a better deep learning model is obtained by integrating several cutting-edge deep learning models. Meanwhile, to make the integrated learning model converge better and faster, the parameters of the integrated module are initialized, constraints are imposed, and a new activation function is designed for better integration of the sub-models. Finally, the integrated large model is distilled for knowledge distillation, which greatly reduces the number of model parameters and improves the model inference efficiency.

Anisotropic Total Variation Regularized Low-Rank Approximation for SSS Images Radiometric Distortion Correction

Abstract: Radiometric distortion caused by the time-varying gain (TVG), beam patterns, angular responses, and sonar altitude variations, highly degrades the quality of side-scan sonar (SSS) images. Thus, radiometric distortion correction becomes a fundamental step for SSS image processing, which holds vital importance for geomorphic applications. However, existing methods cannot take the prior information of the acoustic illumination component as well as the feature of seafloor into consideration well, which would easily cause damage to the image and also always be powerless for residual stripe noise. In this article, a novel radiometric correction method is proposed. First, we give a detailed analysis of the SSS imaging theory based on Lambert’s law as well as prior knowledge about the characteristics of SSS images. Then, incorporating the prior of the SSS imaging process, the low-rank constraint is specifically introduced for the illumination component, while the anisotropic total variation (ATV) constraint is used to constraint the albedo component; combining other constraints, a decomposition model is proposed to correct the radiometric distortion based on the SSS imaging theory. Also, an alternative minimization method has been adopted to solve the proposed model effectively. Experiments proved the validity of the proposed method.