Feature Transformation Research Articles

PDFormer: Propagation Delay-Aware Dynamic Long-Range Transformer for Traffic Flow Prediction

As a core technology of Intelligent Transportation System, traffic flow prediction has a wide range of applications. The fundamental challenge in traffic flow prediction is to effectively model the complex spatial-temporal dependencies in traffic data. Spatial-temporal Graph Neural Network (GNN) models have emerged as one of the most promising methods to solve this problem. However, GNN-based models have three major limitations for traffic prediction: i) Most methods model spatial dependencies in a static manner, which limits the ability to learn dynamic urban traffic patterns; ii) Most methods only consider short-range spatial information and are unable to capture long-range spatial dependencies; iii) These methods ignore the fact that the propagation of traffic conditions between locations has a time delay in traffic systems. To this end, we propose a novel Propagation Delay-aware dynamic long-range transFormer, namely PDFormer, for accurate traffic flow prediction. Specifically, we design a spatial self-attention module to capture the dynamic spatial dependencies. Then, two graph masking matrices are introduced to highlight spatial dependencies from short- and long-range views. Moreover, a traffic delay-aware feature transformation module is proposed to empower PDFormer with the capability of explicitly modeling the time delay of spatial information propagation. Extensive experimental results on six real-world public traffic datasets show that our method can not only achieve state-of-the-art performance but also exhibit competitive computational efficiency. Moreover, we visualize the learned spatial-temporal attention map to make our model highly interpretable.

SEFormer: Structure Embedding Transformer for 3D Object Detection

Effectively preserving and encoding structure features from objects in irregular and sparse LiDAR points is a crucial challenge to 3D object detection on the point cloud. Recently, Transformer has demonstrated promising performance on many 2D and even 3D vision tasks. Compared with the fixed and rigid convolution kernels, the self-attention mechanism in Transformer can adaptively exclude the unrelated or noisy points and is thus suitable for preserving the local spatial structure in the irregular LiDAR point cloud. However, Transformer only performs a simple sum on the point features, based on the self-attention mechanism, and all the points share the same transformation for value. A such isotropic operation cannot capture the direction-distance-oriented local structure, which is essential for 3D object detection. In this work, we propose a Structure-Embedding transFormer (SEFormer), which can not only preserve the local structure as a traditional Transformer but also have the ability to encode the local structure. Compared to the self-attention mechanism in traditional Transformer, SEFormer learns different feature transformations for value points based on the relative directions and distances to the query point. Then we propose a SEFormer-based network for high-performance 3D object detection. Extensive experiments show that the proposed architecture can achieve SOTA results on the Waymo Open Dataset, one of the most significant 3D detection benchmarks for autonomous driving. Specifically, SEFormer achieves 79.02% mAP, which is 1.2% higher than existing works. https://github.com/tdzdog/SEFormer.

Joint Diagnostic Method of Tumor Tissue Based on Hyperspectral Spectral-Spatial Transfer Features

In order to improve the clinical application of hyperspectral technology in the pathological diagnosis of tumor tissue, a joint diagnostic method based on spectral-spatial transfer features was established by simulating the actual clinical diagnosis process and combining micro-hyperspectral imaging with large-scale pathological data. In view of the limited sample volume of medical hyperspectral data, a multi-data transfer model pre-trained on conventional pathology datasets was applied to the classification task of micro-hyperspectral images, to explore the differences in spectral-spatial transfer features in the wavelength of 410–900 nm between tumor tissues and normal tissues. The experimental results show that the spectral-spatial transfer convolutional neural network (SST-CNN) achieved a classification accuracy of 95.46% for the gastric cancer dataset and 95.89% for the thyroid cancer dataset, thus outperforming models trained on single conventional digital pathology and single hyperspectral data. The joint diagnostic method established based on SST-CNN can complete the interpretation of a section of data in 3 min, thus providing a new technical solution for the rapid diagnosis of pathology. This study also explored problems involving the correlation between tumor tissues and typical spectral-spatial features, as well as the efficient transformation of conventional pathological and transfer spectral-spatial features, which solidified the theoretical research on hyperspectral pathological diagnosis.

Folk art of China: artistic characteristics of New Year's paintings "puhui Nianhua" of Gaomi city, Shandong Province

This study is devoted to the artistic and stylistic analysis of traditional Chinese New Year's pictures called "puhui nianhua" on the example of historical and documentary material collected in the city of Gaomi in Shandong province. It traces the changes in its form and content, the means of expression used during the five-hundred-year history of its distribution in China. The subject of attention is the development of Chinese splint pictures from the Ming Dynasty to the present. As an object, we see the transformation of technological and visual features of the performance of such works within their individual varieties, which were popular in the creative activities of several generations of masters of the city of Gaomi. The purpose of the study is to get a general idea of how the main varieties of "puhui nianhua" were born and formed in the creative activity of Gaoming artisan artists. By means of artistic-stylistic, iconographic and comparative-historical analysis, the author identifies several stages in the process of formation of the tradition of making such pictures. Moreover, a typological series of Chinese New Year's splints is proposed, consisting of images of Lao and Buddhist saints, visualizations of homophonic metaphors, and a description of Chinese life. Each of the presented types differs not only in the originality of the content, but also in the set of compositional and visual means used by the authors.

PTC-Net: Point-Wise Transformer With Sparse Convolution Network for Place Recognition

In the point-cloud-based place recognition area, the existing hybrid architectures combining both convolutional networks and transformers have shown promising performance. They mainly apply the voxel-wise transformer after the sparse convolution (SPConv). However, they can induce information loss by the sparse voxelization and further result in loss propagation to the transformer, significantly degrading the performance of the network, especially in outdoor scenes with complex geometric structures and multiple small objects. To address this issue, we propose a novel Point-wise Transformer with sparse Convolution (PTC). Specifically, SPConv is applied to the sparsely voxelized point cloud to extract local features, which are then converted to the point-based representation via a feature transformation unit (FTU). As such, our PTC can apply a transformer model based on the point-wise representation rather than on the voxel-wise one. To enhance the ability to capture long-range features and reduce the computational complexity of the transformer, we propose a two-step transformer, each with different grouping strategies. Meanwhile, in both steps, the attention matrix is computed with much fewer points by grouping a single point cloud into different attention domains. The experiments show that the PTC-Net can achieve state-of-the-art (SOTA) performance, with an improvement of 3.6% on average recall@1. Furthermore, to demonstrate the effectiveness of the PTC, we introduce an extremely light-weight version, PTC-Net-L, with only one PTC layer and half initial channel dimensions, also achieving SOTA performance in terms of the average recall rate and running time with only 0.08 M parameters.

Deep Spatial Feature Transformation for Oriented Aerial Object Detection

Object detection in aerial images has received extensive attention in the field of computer vision. Different from natural images, the aerial objects are usually distributed in any direction. Therefore, the existing detector usually needs more parameters to encode the direction information, resulting in a large number of redundant calculations. In addition, because an ordinary convolution neural network (CNN) does not effectively model the direction change, a large amount of the rotated data is required for the aerial detector. To solve these problems, we propose a Deep Spatial Feature Transformation Network (DSFTNet), which includes a Spatial Feature Extraction Module and a Feature Selection Module. Specifically, we add the rotation convolution kernel to the detector to extract the directional feature of the rotated target to accurately predict the direction of the model. Then we build a dual pyramid to separate the features in the classification and regression tasks. Finally, the polarization function is proposed to construct the critical features that are suitable for their respective tasks, achieving feature selection and more refined detection. Experiments on public remote sensing benchmarks (e.g. DOTA, HRSC2016, and UCASAOD) have proved the effectiveness of our detector.

Preference-Aware Light Graph Convolution Network for Social Recommendation

Social recommendation systems leverage the abundant social information of users existing in the current Internet to mitigate the problem of data sparsity, ultimately enhancing recommendation performance. However, most existing recommendation systems that introduce social information ignore the negative messages passed by high-order neighbor nodes and aggregate messages without filtering, which results in a decline in the performance of the recommendation system. Considering this problem, we propose a novel social recommendation model based on graph neural networks (GNNs) called the preference-aware light graph convolutional network (PLGCN), which contains a subgraph construction module using unsupervised learning to classify users according to their embeddings and then assign users with similar preferences to a subgraph to filter useless or even negative messages from users with different preferences to attain even better recommendation performance. We also designed a feature aggregation module to better combine user embeddings with social and interaction information. In addition, we employ a lightweight GNN framework to aggregate messages from neighbors, removing nonlinear activation and feature transformation operations to alleviate the overfitting problem. Finally, we carried out comprehensive experiments using two publicly available datasets, and the results indicate that PLGCN outperforms the current state-of-the-art (SOTA) method, especially in dealing with the problem of cold start. The proposed model has the potential for practical applications in online recommendation systems, such as e-commerce, social media, and content recommendation.

COVID-19 Media Chatter and Macroeconomic Reflectors on Black Swan: A Spanish and Indian Stock Markets Comparison

Predictive analytics of financial markets in developed and emerging economies during the COVID-19 regime is undeniably challenging due to unavoidable uncertainty and the profound proliferation of negative news on different platforms. Tracking the media echo is crucial to explaining and anticipating the abrupt fluctuations in financial markets. The present research attempts to propound a robust framework capable of channeling macroeconomic reflectors and essential media chatter-linked variables to draw precise forecasts of future figures for Spanish and Indian stock markets. The predictive structure combines Isometric Mapping (ISOMAP), which is a non-linear feature transformation tool, and Gradient Boosting Regression (GBR), which is an ensemble machine learning technique to perform predictive modelling. The Explainable Artificial Intelligence (XAI) is used to interpret the black-box type predictive model to infer meaningful insights. The overall results duly justify the incorporation of local and global media chatter indices in explaining the dynamics of respective financial markets. The findings imply marginally better predictability of Indian stock markets than their Spanish counterparts. The current work strives to compare and contrast the reaction of developed and developing financial markets during the COVID-19 pandemic, which has been argued to share a close resemblance to the Black Swan event when applying a robust research framework. The insights linked to the dependence of stock markets on macroeconomic indicators can be leveraged for policy formulations for augmenting household finance.

Quaternion-based graph convolution network for recommendation

Graph Convolution Network (GCN) has been widely applied in recommender systems for its representation learning capability on user and item embeddings. However, GCN is vulnerable to noisy and incomplete graphs, which are common in real world, due to its recursive message propagation mechanism. In the literature, some work proposes to remove the feature transformation during message propagation, but making it unable to effectively capture the graph structural features. Moreover, they model users and items in the Euclidean space, which has been demonstrated to have high distortion when modeling complex graphs, further degrading the capability to capture the graph structural features and leading to sub-optimal performance. To this end, in this paper, we propose a simple yet effective Quaternion-based Graph Convolution Network (QGCN) recommendation model. In the proposed model, we utilize the hyper-complex Quaternion space to learn user and item representations and feature transformation to improve both performance and robustness. Specifically, we first embed all users and items into the Quaternion space. Then, we introduce the quaternion embedding propagation layers with quaternion feature transformation to perform message propagation. Finally, we combine the embeddings generated at each layer with the mean pooling strategy to obtain the final embeddings for recommendation. Extensive experiments on three public benchmark datasets demonstrate that our proposed QGCN model outperforms baseline methods by a large margin.

EFFC-Net: lightweight fully convolutional neural networks in remote sensing disaster images

ABSTRACT Continuous development of remote sensing technology can rapidly and accurately extract secondary disaster information, such as the area of various disasters. However, in the extraction process, some disasters should be initially classified and identified. In view of this concept, a lightweight fully Convolutional Neural Network (CNN) model Earthquake – Flood–Fire – Cyclone (EFFC)-Net is proposed. Two modules, EFFC_Block and EFFC_Tran_Block, which are used for feature extraction and feature transformation, respectively, are introduced. The EFFC-Net network model is reconstructed through the EFFC_Block and EFFC_Tran_Block modules. Subsequently, EFFC-Net is compared with CNN and transformer models. Results show that the EFFC-Net network model performs effectively in precision, recall, F1_score, and parameters, outperforming the more advanced CNN and transformer models. Moreover, the test time of the Cifar_10 and Cifar_100 datasets was compared, and the results indicate that the EFFC-Net algorithm has the shortest running time and achieves the lightweight goal. Therefore, the EFFC-Net lightweight fully CNN has high disaster classification application value and good portability.

An HMM-DNN-Based System for the Detection and Classification of Low-Frequency Acoustic Signals from Baleen Whales, Earthquakes, and Air Guns off Chile

Marine passive acoustic monitoring can be used to study biological, geophysical, and anthropogenic phenomena in the ocean. The wide range of characteristics from geophysical, biological, and anthropogenic sounds sources makes the simultaneous automatic detection and classification of these sounds a significant challenge. Here, we propose a single Hidden Markov Model-based system with a Deep Neural Network (HMM-DNN) for the detection and classification of low-frequency biological (baleen whales), geophysical (earthquakes), and anthropogenic (air guns) sounds. Acoustic data were obtained from the Preparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization station off Juan Fernandez, Chile (station HA03) and annotated by an analyst (498 h of audio data containing 30,873 events from 19 different classes), and then divided into training (60%), testing (20%), and tuning (20%) subsets. Each audio frame was represented as an observation vector obtained through a filterbank-based spectral feature extraction procedure. The HMM-DNN training procedure was carried out discriminatively by setting HMM states as targets. A model with Gaussian Mixtures Models and HMM (HMM-GMM) was trained to obtain an initial set of HMM target states. Feature transformation based on Linear Discriminant Analysis and Maximum Likelihood Linear Transform was also incorporated. The HMM-DNN system displayed good capacity for correctly detecting and classifying events, with high event-level accuracy (84.46%), high weighted average sensitivity (84.46%), and high weighted average precision (89.54%). Event-level accuracy increased with higher event signal-to-noise ratios. Event-level metrics per class also showed that our HMM-DNN system generalized well for most classes but performances were best for classes that either had a high number of training exemplars (e.g., generally above 50) and/or were for classes of signals that had low variability in spectral features, duration, and energy levels. Fin whale and Antarctic blue whale song and air guns performed particularly well.

Patch merging refiner embedding UNet for image denoising

Image denoising is a fundamental task in low-level computer vision. The well-known CNN-UNet model with encoder and decoder has shown excellent denoising performance. The encoder adopts downsample modules such as pooling, convolution and patch merging to reduce the feature map size. The downsample module is very important for feature transformation between adjacent feature layers. However, these plain downsample modules are incapable of removing noise during feature transformation. It is expected that the downsample module can play the role of removing noise while executing feature transformation. Therefore, we design a patch merging refiner (PMR) downsample module, which utilizes subspace projection to learn a set of restoration basis from the feature space and projects the patch merging feature onto such space, to remove noise and retain the authentic information of feature space while executing feature transformation. Subsequently, from the perspective of the non-local mechanism, we carry out group convolution (GC) block module on PMR to restore high-frequency detail. Finally, by integrating the two modules into the commonly-used UNet model, we construct a PMR-UNet architecture. Experiments on synthetic and real-world noise images demonstrate that our architecture has stronger anti-noise power and surpasses the state-of-the-art denoisers in both quantitative measure and visual perception.

Application of Artificial Intelligence for Fraudulent Banking Operations Recognition

This study considers the task of applying artificial intelligence to recognize bank fraud. In recent years, due to the COVID-19 pandemic, bank fraud has become even more common due to the massive transition of many operations to online platforms and the creation of many charitable funds that criminals can use to deceive users. The present work focuses on machine learning algorithms as a tool well suited for analyzing and recognizing online banking transactions. The study’s scientific novelty is the development of machine learning models for identifying fraudulent banking transactions and techniques for preprocessing bank data for further comparison and selection of the best results. This paper also details various methods for improving detection accuracy, i.e., handling highly imbalanced datasets, feature transformation, and feature engineering. The proposed model, which is based on an artificial neural network, effectively improves the accuracy of fraudulent transaction detection. The results of the different algorithms are visualized, and the logistic regression algorithm performs the best, with an output AUC value of approximately 0.946. The stacked generalization shows a better AUC of 0.954. The recognition of banking fraud using artificial intelligence algorithms is a topical issue in our digital society.

Self-supervised Dynamic Graph Embedding with evolutionary neighborhood and community

Real-world data and systems are generally evolutionary over time, which can be formalized as dynamic graphs, and Dynamic Graph Embedding (DGE) as a fundamental technique for analyzing dynamic graphs, has shown considerable potential in recent years. However, the majority of existing DGEs focus only on the local neighborhood structure, ignoring the global community structure, which is also an important property of dynamic graphs. Therefore, how to fuse both local and global information to learn more comprehensive graph representations and their variation patterns is a meaningful and promising topic. In this paper, we propose a novel DGE method to explore the graph properties and evolution patterns from Neighborhood and Community views, named DGNC. The neighborhood view learns the node embeddings in line with timestamped interaction events, and models the local variation of graphs via the time-varying attention vector. While the community view encodes the alternative transformation of node features and community features according to community correlation, which aims to extract the global evolution of graphs. To better integrate the two views, cross-supervised contrastive learning is introduced to enable them to collaboratively provide guidance and complementary information to each other, which can significantly improve the quality of graph embedding. We evaluate DGNC through diverse downstream tasks on four real-world dynamic graph datasets, and the experimental results verify the effectiveness and superiority of our proposed method.

Accurate compressed traffic detection via traffic analysis using Graph Convolutional Network based on graph structure feature

As the application of data compression technology expands in areas such as IoT, webpage, and video data transmission, there are problems such as leakage of compressed but unencrypted user data, difficulty in supervising compressed data, and confusion between compressed and private encrypted traffic. Existing compressed traffic detection methods are significantly affected by data length and rely on binary classification in one step using random tests. It remains a challenging task to conduct accurate and efficient compressed traffic detection. In this paper, we present GCN-RTG, a compressed traffic detection method using Graph Convolutional Network. We investigate the randomness feature transformation pattern of packet sequences, propose a graph structure based on this pattern, and design a powerful GCN-based classifier to detect compressed traffic. The experimental results show that GCN-RTG achieves 94% accuracy in compressed traffic detection, remarkably improving by nearly 10% accuracy compared with traditional machine learning methods and approximately 5% compared with CNN and LSTM. Considering the effect of private encrypted traffic, GCN-RTG attains an accuracy of 89% for detecting compressed traffic. Furthermore, GCN-RTG can maintain an 83% accuracy even in the most extreme packet loss scenario and reach an outstanding accuracy of up to 95% of compressed traffic detection in real-world network data sized 4GB from the Jiangsu education backbone network in China.

A research study on new energy brand users based on principal component analysis (PCA) and fusion target planning model for sustainable environment of smart cities

The preservation of green environment is inevitable in order to reduce the carbon emissions from the atmosphere. The development of infrastructure, vehicle technology and global business opportunities are giving rise to carbon footprints. The production of energy based vehicles is the key for all the nations to reduce the carbon emissions to a great extent and preserve the green environment. The research in the area of energy conservation is promoted world-wide by switching from patrol/diesel based engines to electric vehicles. The consistent efforts are made by many researchers and industries to save energy by consuming minimal electricity to drive the vehicle. There is also a need to create awareness among the users of electric vehicles that their effort to use electric vehicles may help the mother earth from the further deterioration and in reduction of air pollution. Without creating awareness among the users, the initiative of green planet cannot be completed successfully. In order to prove the hypothesis of the research study, the data is collected through a survey which is based on the questionnaire on energy vehicle consumers. This research study analyzes the potential users of energy based vehicles and their expectations from the energy based vehicles. The feature dimensions of the data source are finalized and the size range of the value is determined. Then the feature transformations is devised such as feature reduction and feature merger to make changes in the feature dimension of the original data source to facilitate the solution of the problem statement. Then the statistical analysis is performed with the aid of detection of outliers and missing values, interpretation of data, correlation between the variables and analysis of data. This model can be adopted by the manufacturers of electric vehicles to create awareness in the users about the benefits of using electric vehicles such as reduction in carbon emissions, improvement in air quality index (AQI), saving of natural resources and protection of green environment.

Comparative Analysis of Machine and Deep Learning Models for Soil Properties Prediction from Hyperspectral Visual Band

Estimating various properties of soil, including moisture, carbon, and nitrogen, is crucial for studying their correlation with plant health and food production. However, conventional methods such as oven-drying and chemical analysis are laborious, expensive, and only feasible for a limited land area. With the advent of remote sensing technologies like multi/hyperspectral imaging, it is now possible to predict soil properties non-invasive and cost-effectively for a large expanse of bare land. Recent research shows the possibility of predicting those soil contents from a wide range of hyperspectral data using good prediction algorithms. However, these kinds of hyperspectral sensors are expensive and not widely available. Therefore, this paper investigates different machine and deep learning techniques to predict soil nutrient properties using only the red (R), green (G), and blue (B) bands data to propose a suitable machine/deep learning model that can be used as a rapid soil test. Another objective of this research is to observe and compare the prediction accuracy in three cases i. hyperspectral band ii. full spectrum of the visual band, and iii. three-channel of RGB band and provide a guideline to the user on which spectrum information they should use to predict those soil properties. The outcome of this research helps to develop a mobile application that is easy to use for a quick soil test. This research also explores learning-based algorithms with significant feature combinations and their performance comparisons in predicting soil properties from visual band data. For this, we also explore the impact of dimensional reduction (i.e., principal component analysis) and transformations (i.e., empirical mode decomposition) of features. The results show that the proposed model can comparably predict the soil contents from the three-channel RGB data.

An Integrated Statistical and Clinically Applicable Machine Learning Framework for the Detection of Autism Spectrum Disorder

Autism Spectrum Disorder (ASD) is a neurological impairment condition that severely impairs cognitive, linguistic, object recognition, interpersonal, and communication skills. Its main cause is genetic, and early treatment and identification can reduce the patient’s expensive medical costs and lengthy examinations. We developed a machine learning (ML) architecture that is capable of effectively analysing autistic children’s datasets and accurately classifying and identifying ASD traits. We considered the ASD screening dataset of toddlers in this study. We utilised the SMOTE method to balance the dataset, followed by feature transformation and selection methods. Then, we utilised several classification techniques in conjunction with a hyperparameter optimisation approach. The AdaBoost method yielded the best results among the classifiers. We employed ML and statistical approaches to identify the most crucial characteristics for the rapid recognition of ASD patients. We believe our proposed framework could be useful for early diagnosis and helpful for clinicians.

Cross-stream contrastive learning for self-supervised skeleton-based action recognition

Self-supervised skeleton-based action recognition enjoys a rapid growth along with the development of contrastive learning. The existing methods rely on imposing invariance to augmentations of 3D skeleton within a single data stream, which merely leverages the easy positive pairs and limits the ability to explore the complicated movement patterns. In this paper, we advocate that the defect of single-stream contrast and the lack of necessary feature transformation are responsible for easy positives, and therefore propose a Cross-Stream Contrastive Learning framework for skeleton-based action Representation learning (CSCLR). Specifically, the proposed CSCLR not only utilizes intra-stream contrast pairs, but introduces inter-stream contrast pairs as hard samples to formulate a better representation learning. Besides, to further exploit the potential of positive pairs and increase the robustness of self-supervised representation learning, we propose a Positive Feature Transformation (PFT) strategy which adopts feature-level manipulation to increase the variance of positive pairs. To validate the effectiveness of our method, we conduct extensive experiments on three benchmark datasets NTU-RGB + D 60, NTU-RGB + D 120 and PKU-MMD. Experimental results show that our proposed CSCLR exceeds the state-of-the-art methods on a diverse range of evaluation protocols.

Colorectal polyp segmentation method based on fusion of transformer and cross-level phase awareness

In order to address the issues of spatial induction bias and lack of effective representation of global contextual information in colon polyp image segmentation, which lead to the loss of edge details and mis-segmentation of lesion areas, a colon polyp segmentation method that combines Transformer and cross-level phase-awareness is proposed. The method started from the perspective of global feature transformation, and used a hierarchical Transformer encoder to extract semantic information and spatial details of lesion areas layer by layer. Secondly, a phase-aware fusion module (PAFM) was designed to capture cross-level interaction information and effectively aggregate multi-scale contextual information. Thirdly, a position oriented functional module (POF) was designed to effectively integrate global and local feature information, fill in semantic gaps, and suppress background noise. Fourthly, a residual axis reverse attention module (RA-IA) was used to improve the network's ability to recognize edge pixels. The proposed method was experimentally tested on public datasets CVC-ClinicDB, Kvasir, CVC-ColonDB, and EITS, with Dice similarity coefficients of 94.04%, 92.04%, 80.78%, and 76.80%, respectively, and mean intersection over union of 89.31%, 86.81%, 73.55%, and 69.10%, respectively. The simulation experimental results show that the proposed method can effectively segment colon polyp images, providing a new window for the diagnosis of colon polyps.