Multi-directional Feature Research Articles

Revisiting non-learned operators based deep learning for image classification: a lightweight directional-aware network

Due to the stable feature representation capability provided by non-learned operators, the integration with deep learning models, i.e., non-learned operator based deep learning models, has become a paradigm, however, performance-wise, it is still not promising. In this paper, by revisiting non-learned operator based deep learning models, we reveal the reasons for their underperformance: lack of geometric invariance, insufficient sparsity, and neglect of directional importance. In response, we present a Lightweight Directional-Aware Network (LDAN) for image classification. Specifically, to generate sparse geometric-invariant features, we propose a ShearletNet to capture multi-directional features in three different levels. Then, a Directional-Aware module is designed to highlight the discriminative multi-directional features and generate multi-scale features. Finally, a Pointwise Convolution module is used to integrate the multi-directional features with the multi-scale ones for reducing the computational resources. Experiments on the commonly used CIFAR10, CIFAR100, Self-Taught Learning 10 (STL10), and Tiny ImageNet datasets demonstrate the efficiency and effectiveness of the proposed LDAN. Compared to the existing non-learned operator based models, LDAN reduces the parameter count by 80.83% while achieving a 6.32% increase in accuracy.

Map-Matching Algorithm Based on Hidden Markov and Constraint Value Pruning

Map matching is the process of matching global positioning system (GPS) trajectory data with map data. Its purpose is to determine the actual route of the moving object. Because of factors such as positioning devices and the environment, the GPS trajectory data obtained may not be accurate. In location-based services, map matching can be used to address the accuracy and reliability issues of GPS location data. There are currently many map-matching algorithms, for example, spatio-temporal-based (STD) matching algorithm, improved interactive voting-based map-matching (IIVMM) algorithm, and turning-point-based (TPB) offline map-matching algorithm, but existing algorithms have some shortcomings, such as low matching accuracy on complex roads or low sampling rates, and routing calculation time bottlenecks. Therefore, this paper proposes a fast matching algorithm based on constrained value pruning that is suitable for complex roads. The algorithm considers the multi-directional features within the road and uses secondary calculations to determine candidate points, improving the accuracy of candidate point selection. Additionally, two pruning strategies based on constrained values are introduced to reduce the majority of the routing calculation process and improve the matching efficiency. Finally, comparative experiments are conducted on a real trajectory dataset. The results show that, compared with STD, IIVMM, and TPB algorithms, the algorithm accuracy is improved by about 2% to 4%, and the running time is reduced by about 30%.

Multiresolution Interpretable Contourlet Graph Network for Image Classification.

Modeling contextual relationships in images as graph inference is an interesting and promising research topic. However, existing approaches only perform graph modeling of entities, ignoring the intrinsic geometric features of images. To overcome this problem, a novel multiresolution interpretable contourlet graph network (MICGNet) is proposed in this article. MICGNet delicately balances graph representation learning with the multiscale and multidirectional features of images, where contourlet is used to capture the hyperplanar directional singularities of images and multilevel sparse contourlet coefficients are encoded into graph for further graph representation learning. This process provides interpretable theoretical support for optimizing the model structure. Specifically, first, the superpixel-based region graph is constructed. Then, the region graph is applied to code the nonsubsampled contourlet transform (NSCT) coefficients of the image, which are considered as node features. Considering the statistical properties of the NSCT coefficients, we calculate the node similarity, i.e., the adjacency matrix, using Mahalanobis distance. Next, graph convolutional networks (GCNs) are employed to further learn more abstract multilevel NSCT-enhanced graph representations. Finally, the learnable graph assignment matrix is designed to get the geometric association representations, which accomplish the assignment of graph representations to grid feature maps. We conduct comparative experiments on six publicly available datasets, and the experimental analysis shows that MICGNet is significantly more effective and efficient than other algorithms of recent years.

Global induced local network for infrared: dim small target detection

Abstract It is challenging to detect infrared dim targets submerged in complicated backgrounds due to their small size and faint intensity. The previous attention-based detection networks frequently require global long-range dependence. Significant calculations are required to determine the target's sparse but meaningful position. To prevent wasting calculations on the background, this paper offers a detection network guided by global context for local feature learning, named GILNet (Global Induced Local Network). It designs a global location module (GLM) and a local feature interaction module (LFIM) to capture the global position and features of targets, respectively. More specifically, using global context interaction, the GLM finds the region that might contain dim small targets, that is, the coarse location. In the coarsely located regions, the LFIM further acquires feature information about targets. Next, we also design an eight-directional attention operation to obtain the contour information of targets in the low feature map. It is fused with the high feature map in the multi-directional feature fusion module (MFFM), which retains more semantic and spatial information about targets. Finally, quantitative and qualitative analysis show that the GILNet performs better than eight comparison methods on two public datasets.

Multidirectional Attention Fusion Network for SAR Change Detection

Synthetic Aperture Radar (SAR) imaging is essential for monitoring geomorphic changes, urban transformations, and natural disasters. However, the inherent complexities of SAR, particularly pronounced speckle noise, often lead to numerous false detections. To address these challenges, we propose the Multidirectional Attention Fusion Network (MDAF-Net), an advanced framework that significantly enhances image quality and detection accuracy. Firstly, we introduce the Multidirectional Filter (MF), which employs side-window filtering techniques and eight directional filters. This approach supports multidirectional image processing, effectively suppressing speckle noise and precisely preserving edge details. By utilizing deep neural network components, such as average pooling, the MF dynamically adapts to different noise patterns and textures, thereby enhancing image clarity and contrast. Building on this innovation, MDAF-Net integrates multidirectional feature learning with a multiscale self-attention mechanism. This design utilizes local edge information for robust noise suppression and combines global and local contextual data, enhancing the model’s contextual understanding and adaptability across various scenarios. Rigorous testing on six SAR datasets demonstrated that MDAF-Net achieves superior detection accuracy compared with other methods. On average, the Kappa coefficient improved by approximately 1.14%, substantially reducing errors and enhancing change detection precision.

Multi-directional feature fusion super-resolution network based on nonlinear spiking neural P systems

Single-image super-resolution (SISR) aims to recover lost details and textures from limited information to improve the visual quality and detail clarity of an image. Most of the existing convolutional neural network (CNN)-based SISR methods do not make full use of the intermediate features of the network, and at the same time its processing process can result in the loss of information details. To address these problems, this paper recommends a novel multi-directional feature fusion super-resolution network. The direction-aware mechanism with nonlinear spiking neural P (NSNP) systems is integrated to design a multidirectional feature extraction module, while different NSNP-like convolution modules are constructed for extracting deep features on the basis of nonlinear spiking mechanism. Moreover, in the residual feature enhancement module, three different branches are designed using these NSNP-like convolution modules for feature enhancement and fusion to decrease the loss of information. The preposed network is evaluated on five benchmark datasets, and the experimental results show that the network performs a better reconstruction with good capability of generalization compared to most state-of-the-art methods.

An effective method for small objects detection based on MDFFAM and LKSPP

Object detection is one of the research hotspots in computer vision. However, most existing object detectors struggle with the identification of small targets. Therefore, the paper proposes two modules: the MDFFAM (Multi-Directional Feature Fusion Attention Mechanism) and the LKSPP (Large Kernel Spatial Pyramid Pooling), to enhance the detector's effectiveness in identifying subtle faults on the surface of mechanical equipment. LKSPP aims to expand the receptive field to capture high-level semantic features through large kernels. Meanwhile, the MDFFAM allows the network to efficiently utilize spatial location information and adaptively recognize detection priorities. In the detection task, MDFFAM effectively captures feature information in three spatial directions: width, height, and channel, with the location information fully utilized to establish stable long-range dependencies. Moreover, LKSPP boasts a larger receptive field and imposes less computational burden compared to the SPPCSPC by YOLOv7. Finally, experiments demonstrate that the proposed module effectively improves the detection accuracy for small targets, surpassing the state-of-the-art object detector, YOLOv7. Remarkably, MDFFAM incurs almost negligible computational overhead.

Robust Coarse-to-Fine Registration Algorithm for Optical and SAR Images Based on Two Novel Multiscale and Multidirectional Features

Robust Coarse-to-Fine Registration Algorithm for Optical and SAR Images Based on Two Novel Multiscale and Multidirectional Features

A Multiscale and Multidirection Feature Fusion Network for Road Detection From Satellite Imagery

A Multiscale and Multidirection Feature Fusion Network for Road Detection From Satellite Imagery

Multiscale and Multidirection Feature Extraction Network for Hyperspectral and LiDAR Classification

Multiscale and Multidirection Feature Extraction Network for Hyperspectral and LiDAR Classification

A label-relevance multi-direction interaction network with enhanced deformable convolution for forest smoke recognition

Forest fires pose a significant threat to both the economy and ecology, causing extensive damage. Smoke serves as a crucial indicator of forest fires, often appearing before the actual fire. However, existing methods for smoke recognition are susceptible to missed alarms and false alarms due to the diverse nature of smoke and the presence of various continuous interferences in complex real-world scenarios. To tackle these challenges, this paper proposes a label-relevance multi-direction interaction network with enhanced deformable convolution. Firstly, to ensure the extraction of robust features, we propose an enhanced deformable convolution module that breaks away from fixed geometric structures and incorporates interval information. Secondly, to prevent high-response high-level features from overshadowing low-level features during the feature interaction process, we introduce a multi-directional feature interaction module that obtains complementary features from different convolution layers. Lastly, to leverage the relevance and pixel distribution information in the label image, we propose a new loss term based on generative adversarial learning. This loss term measures the distribution similarity between the network's predictions and the ground truth. Through extensive experiments, we demonstrate that our model accurately estimates smoke pixels and outperforms existing smoke recognition methods.

Local Binary Convolution Based Prior Knowledge of Multi-Direction Features for Finger Vein Verification

Local Binary Convolution Based Prior Knowledge of Multi-Direction Features for Finger Vein Verification

Two-stage convolutional neural network for joint removal of sensor noise and background interference in lock-in thermography

Lock-in thermography provides a fast, low-cost, and accurate non-destructive testing for subsurface defect detection in materials. However, there is serious sensor noise and background interference presented in the infrared images captured by lock-in thermography. In this paper, we set up lock-in thermography experiments to analyze the characteristics of sensor noise and background interference. Based on their different characteristics, we design a two-stage CNN model for the joint removal of sensor noise and background interference. This model consists of a number of multi-direction feature extraction blocks (MDFEBs) to extract local features in different directions for pixel-level sensor noise removal, and an encoder–decoder architecture with large receptive fields to extract global features for background interference correction. Experimental results demonstrate that our proposed method outperforms the well-established infrared image denoising and non-uniform background correction techniques on specimens made of different materials and at various excitation frequencies. The proposed two-stage CNN model provides high-quality input of infrared images for the following subsurface defect detection task in lock-in thermography.

Table Tennis Player Expression Recognition Method Based on Gabor Multi directional Feature Fusion

Table Tennis Player Expression Recognition Method Based on Gabor Multi directional Feature Fusion

Robust Face Recognition Algorithm Based on Multidirectional Log-Gabor Features

In order to improve the robustness to variable factors such as illumination, expression and pose, a face recognition method based on multi-directional Log-Gabor feature map sparse representation is proposed. First, conduct multi-directional and multi-scale Log-Gabor transformation on the face image, and then fuse the Log-Gabor features of different scales in the same direction to obtain a multi-directional feature map. Then, extract Gist features from the fused feature map of each direction and give them adaptive weights. Then, connect the adaptive weighted Gist features of all directional feature maps to form a feature vector of the face image, Finally, sparse representation classification method is used to realize face recognition. Face recognition experiments were carried out on WIDER FACE dataset and the MAFA occlusion face dataset, and the identification accuracy of the proposed method is compared with SRC, NMR, Gabor NLDA NN and Gabor NLDA SVM. The experimental results show that the method in this paper can effectively describe the local face information by using multi-directional Log Gabor feature map (MGFM) to establish the feature representation of the face image, and the improved cooperative representation classification algorithm using adaptive K nearest neighbor strategy can achieve higher recognition accuracy.

A new condition-monitoring method based on multi-variable correlation learning network for wind turbine fault detection

Reasonable and in-depth analysis of the supervisory control and data acquisition (SCADA) dataset can improve the accuracy and reliability of anomaly detection in wind turbines. In this paper, a multi-variable correlation learning network named the attention mechanism temporal convolutional network–gated recurrent unit (AMTCN-GRU) is proposed to extract the multidirectional features of SCADA data for wind turbine condition monitoring. First, the parameters with greater relevance to the prediction target are selected as input parameters of this method. Meanwhile, the cabin vibration signal contains the transient characteristics of the operating system. If the component connected to the cabin fails, the vibration signal will change immediately. Then, the vibration parameter is selected as one of the inputs. In this paper, a novel AMTCN model is proposed to enhance the feature extraction capability, which is constituted by a convolutional block attention mechanism embedded to the TCN’s residual block structure. The extracted features can be weighted again to make the output more relevant to the predicted target. GRU is performed to construct the connections of feature and output for the condition prediction of the wind turbine. Finally, it is proven that the proposed method can accurately and reliably realize anomaly detection in wind turbines by analyzing the SCADA data of the actual wind farm.

SWASTi-SW: Space Weather Adaptive Simulation Framework for Solar Wind and Its Relevance to the Aditya-L1 Mission

Solar wind streams, acting as a background, govern the propagation of space weather drivers in the heliosphere, which induce geomagnetic storm activities. Therefore, predictions of the solar wind parameters are the core of space weather forecasts. This work presents an indigenous three-dimensional (3D) solar wind model (SWASTi-SW). This numerical framework for forecasting the ambient solar wind is based on a well-established scheme that uses a semiempirical coronal model and a physics-based inner heliospheric model. This study demonstrates a more generalized version of the Wang–Sheeley–Arge relation, which provides a speed profile input to the heliospheric domain. Line-of-sight observations of GONG and Helioseismic and Magnetic Imager magnetograms are used as inputs for the coronal model, which in turn provides the solar wind plasma properties at 0.1 au. These results are then used as an initial boundary condition for the magnetohydrodynamics model of the inner heliosphere to compute the solar wind properties up to 2.1 au. Along with the validation run for multiple Carrington rotations, the effect of variation of specific heat ratio and study of the stream interaction region (SIR) are also presented. This work showcases the multidirectional features of SIRs and provides synthetic measurements for potential observations from the Solar Wind Ion Spectrometer subsystem of the Aditya Solar wind Particle Experiment payload on board ISRO’s upcoming solar mission Aditya-L1.

CSGNet: Cascade semantic guided net for retinal vessel segmentation

Retinal vessels are essential biomarkers for the early diagnosis of ophthalmic diseases. Accurate segmentation of retinal vessels is necessary for further quantitative analysis of fundus images. With the development of deep learning, many deep segmentation models have been proposed and show promising results. Nevertheless, segmentation performance is not satisfactory in tiny vessels and low-contrast vessels. A cascade semantic guided network (CSGNet) is proposed to address this problem. CSGNet follows a low-to-high pipeline, where low-resolution semantic-rich features are first learned. Then, higher-resolution features are learned under the guidance of low-resolution features. In this way, CSGNet is expected to learn both semantic-rich and detail-preserved high-resolution representations to facilitate the segmentation of thin vessels. Meantime, a multi-scale and multi-directional feature learning (M2FL) module is developed, which leverages two strip convolutions and two dilated convolutions to align with the shape of vessels and encode context information. Extensive experiments are conducted to validate the segmentation performance of the proposed method over six public datasets, including four photography datasets (CHASE_DB1, DRIVE, STARE, HRF) and two color scanning laser ophthalmoscopy datasets (IOSTAR, RC-SLO). Also, the proposed method is compared with several popular deep learning-based models. Experimental results show that the proposed method achieves comparable or superior segmentation performance compared with other deep learning-based models with fewer parameters and faster segmentation speed. Meantime, cross-training experiments demonstrate the competitive generalization capability of the proposed method.

The difference in fatigue crack growth induced by internal and external hydrogen in selective laser melted 304L stainless steel

The mechanical properties of selective-laser-melted (SLM) 304L steel under hydrogen charging and hydrogen environment were investigated by tensile and fatigue crack growth (FCG) tests. SLM samples showed higher mechanical properties and weaker hydrogen embrittlement compared with wrought samples, and the FCG rates of SLM 304L tested in a hydrogen environment were higher than hydrogen-charged samples. More obvious cleavage platforms and lamellar structures appeared and cracking tended to the multi-directional feature in the hydrogen gas environment. External hydrogen promotes slip planarity, deformation twinning, and plastic localization near the crack tip, and then causes the acceleration of FCG in SLM 304L steel.

A Hybrid System for Handwritten Character Recognition with High Robustness

In the past few decades, the offline recognition of handwritten Indic scripts has received much attention of researchers. Although an intensive research has been reported for various Indic languages, limited research work is carried out for handwritten Odia, Bangla character recognition owing to their complex shapes and the unavailability of the standard datasets. This paper proposes an automated model for recognizing both handwritten of Odia characters and numerals, along with Bangla numerals with maximum optimization efficiency. The proposed model primarily deals with feature optimization parameters which mainly comprises of three parts. Firstly, the fast discrete curvelet transform (FDCT) is used to derive multidirectional features from the character images. Secondly, PCA along with LDA is used to reduce the dimension of the feature vector. The features are finally subjected to both least-squares support vector machine (LS-SVM), and random forest (RF) for classification. The effectiveness of proposed model is evaluated over three benchmark datasets such as Odia handwritten character, Bangla numeral and Odia handwritten numeral. The efficacy of proposed model achieves superiority as compared to state-of-art techniques. The discriminatory prospective of FDCT along with PCA and LDA features is establish more suitable than its counterparts.