Augmentation Algorithm Research Articles

Chinese Clinical Named Entity Recognition With Segmentation Synonym Sentence Synthesis Mechanism: Algorithm Development and Validation.

Clinical named entity recognition (CNER) is a fundamental task in natural language processing used to extract named entities from electronic medical record texts. In recent years, with the continuous development of machine learning, deep learning models have replaced traditional machine learning and template-based methods, becoming widely applied in the CNER field. However, due to the complexity of clinical texts, the diversity and large quantity of named entity types, and the unclear boundaries between different entities, existing advanced methods rely to some extent on annotated databases and the scale of embedded dictionaries. This study aims to address the issues of data scarcity and labeling difficulties in CNER tasks by proposing a dataset augmentation algorithm based on proximity word calculation. We propose a Segmentation Synonym Sentence Synthesis (SSSS) algorithm based on neighboring vocabulary, which leverages existing public knowledge without the need for manual expansion of specialized domain dictionaries. Through lexical segmentation, the algorithm replaces new synonymous vocabulary by recombining from vast natural language data, achieving nearby expansion expressions of the dataset. We applied the SSSS algorithm to the Robustly Optimized Bidirectional Encoder Representations from Transformers Pretraining Approach (RoBERTa) + conditional random field (CRF) and RoBERTa + Bidirectional Long Short-Term Memory (BiLSTM) + CRF models and evaluated our models (SSSS + RoBERTa + CRF; SSSS + RoBERTa + BiLSTM + CRF) on the China Conference on Knowledge Graph and Semantic Computing (CCKS) 2017 and 2019 datasets. Our experiments demonstrated that the models SSSS + RoBERTa + CRF and SSSS + RoBERTa + BiLSTM + CRF achieved F1-scores of 91.30% and 91.35% on the CCKS-2017 dataset, respectively. They also achieved F1-scores of 83.21% and 83.01% on the CCKS-2019 dataset, respectively. The experimental results indicated that our proposed method successfully expanded the dataset and remarkably improved the performance of the model, effectively addressing the challenges of data acquisition, annotation difficulties, and insufficient model generalization performance.

The Application of Deep Learning in Sports Training

In response to the problems of overfitting, susceptibility to interference information, and insufficient feature expression ability in existing deep learning methods for sports action recognition, the author proposes a deep learning sports action recognition method that integrates attention mechanism. This method proposes a video data augmentation algorithm in data preprocessing to reduce the risk of model overfitting. Then, during the video frame sampling process, the existing sampling algorithms are improved to effectively suppress the influence of interference information. In the special section, the network residual consolidation is proposed to improve the feature extraction capacity of the structure. The Long Term Time Transform (LSTM) network is used to solve the problem of the global correlation of the spatial correlation, and the classification algorithm is achieved by Softmax. and the classification algorithm is proposed. The experimental results show that the recognition rates of this method on UCFYouTube, KTH, and HMDB-51 data are 96.73%, 98.07%, and 64.82%, respectively.

Back analysis of geomechanical parameters based on a data augmentation algorithm and machine learning technique

Accurate geomechanical parameters are key factors for stability evaluation, disaster forecasting, structural design, and supporting optimization. The intelligent back analysis method based on the monitored information is widely recognized as the most efficient and cost-effective technique for inverting parameters. To address the low accuracy of measured data, and the scarcity of comprehensive datasets, this study proposes an innovative back analysis framework tailored for small sample sizes. We introduce a multi-faceted back analysis approach that combines data augmentation with advanced optimization and machine learning techniques. The auxiliary classifier generative adversarial network (ACGAN)-based data augmentation algorithm is first employed to generate synthetic yet realistic samples that adhere to the underlying probability distribution of the original data, thereby expanding the dataset and mitigating the impact of small sample sizes. Subsequently, we harness the power of optimized particle swarm optimization (OPSO) integrated with support vector machine (SVM) to mine the intricate nonlinear relationships between input and output variables. Then, relying on a case study, the validity of the augmented data and the performance of the developed OPSO-SVM algorithms based on two different sample sizes are studied. Results show that the new datasets generated by ACGAN almost coincide with the actual monitored convergences, exhibiting a correlation coefficient exceeding 0.86. Furthermore, the superiority of the OPSO-SVM algorithm is also demonstrated by comparing the displacement prediction capability of various algorithms through four indices. It is also indicated that the relative error of the predicted displacement values reduces from almost 20% to 5% for the OPSO-SVM model trained with 25 samples and that trained with 625 samples. Finally, the inversed parameters and corresponding convergences predicted by the two OPSO-SVM models trained with different samples are discussed, indicating the feasibility of the combination application of ACGAN and OPSO-SVM in back analysis of geomechanical parameters. This endeavor not only facilitates the progression of underground engineering analysis in scenarios with limited data, but also serves as a pivotal reference for both researchers and practitioners alike.

Development of an Algorithm for Semantic Segmentation of Earth Remote Sensing Data to Determine Phytoplankton Populations

Introduction. Computer vision is widely used for semantic segmentation of Earth remote sensing (ERS) data. The method allows monitoring ecosystems, including aquatic ones. Algorithms that maintain the quality of semantic segmentation of ERS images are in demand, specifically, to identify areas with phytoplankton, where water blooms— the cause of suffocation — are possible. The objective of the study is to create an algorithm that processes satellite data as input information for the formation and checking of mathematical models of hydrodynamics, which are used to monitor the state of water bodies. Various algorithms for semantic segmentation are described in the literature. New research focuses on enhancing the reliability of recognition — often using neural networks. This approach is modified in the presented work. To develop the direction, a new set of information from open sources and synthetic data are proposed. They are aimed at improving the generalization ability of the model. For the first time, the contour area of the phytoplankton population is compared to the database — and thus the boundary conditions are formed for the implementation of mathematical models and the construction of boundary-adaptive grids.Materials and Methods. The set of remote sensing images was supplemented with the author's augmentation algorithm in Python. Computer vision segmented areas of phytoplankton populations in the images. The U-Net convolutional neural network (CNN) was trained on the basis of NVIDIA Tesla T4 computing accelerators.Results. To automate the detection of phytoplankton distribution areas, a computer vision algorithm based on the U-Net CNN was developed. The model was evaluated by the calculated values of the main quality metrics related to segmentation tasks. The following metric values were obtained: Precision = 0.89, Recall = 0.88, F1 = 0.87, Dice = 0.87, and IoU = 0.79. Graphical visualization of the results of CNN learning on the training and validation sets showed good quality of model learning. This is evidenced by small changes in the loss function at the end of training. The segmentation performed by the model turned out to be close to manual marking, which indicated the high quality of the proposed solution. The area of the segmented region of the phytoplankton population was calculated by the area of one pixel. The result obtained for the original image was 51202.5 (based on information about the number of pixels related to the bloom of blue-green algae). The corresponding result of the modeling was 51312.Discussion and Conclusion. The study expands theoretical and practical knowledge on the use of convolutional neural networks for semantic segmentation of space imagery data. Given the results of the work, it is possible to assess the potential for automating the process of semantic segmentation of remote sensing data to determine the boundaries of phytoplankton populations using artificial intelligence. The use of the proposed computer vision model to obtain contours of water bloom due to phytoplankton will provide for the creation of databases — the basis for environmental monitoring of water resources and predictive modeling of hydrobiological processes.

A novel load allocation strategy based on the adaptive chiller model with data augmentation

Model-based load allocation strategy is an impactful solution to enhance energy efficiency of multiple-chiller system. Its performance is heavily dependent on the accuracy of chiller model. Data-driven model is a pretty-good solution. However, in real multiple-chiller system, the range of operation condition in historical data is commonly narrow, so it is challenging to develop an accurate data-driven model of chiller throughout full range of operation condition. In this paper, data augmentation algorithm is presented to generate the data outside of historical data, which is based on conditional generative adversarial network (CGAN) and elastic weight consolidation algorithm (EWC). Combined historical data and generated data, augmented training dataset is set up and updated by online operation data. Trained by online updated augmented training dataset periodically, adaptive chiller model is set up. Based on adaptive chiller model, a novel load allocation strategy presented for multiple-chiller system. The proposed strategy is verified by field test in multiple-chiller system. The results show that adaptive chiller model, with the aid of data augmentation algorithm, is more accurate. The proposed strategy can achieve 5.03 % energy saving compared with fixed set-point strategy, and the EER of proposed strategy is 6.27 % higher than that of fixed set-point strategy.

Internal Thread Defect Generation Algorithm and Detection System Based on Generative Adversarial Networks and You Only Look Once.

In the field of industrial inspection, accurate detection of thread quality is crucial for ensuring mechanical performance. Existing machine-vision-based methods for internal thread defect detection often face challenges in efficient detection and sufficient model training samples due to the influence of mechanical geometric features. This paper introduces a novel image acquisition structure, proposes a data augmentation algorithm based on Generative Adversarial Networks (GANs) to effectively construct high-quality training sets, and employs a YOLO algorithm to achieve internal thread defect detection. Through multi-metric evaluation and comparison with external threads, high-similarity internal thread image generation is achieved. The detection accuracy for internal and external threads reached 94.27% and 93.92%, respectively, effectively detecting internal thread defects.

DFLM-YOLO: A Lightweight YOLO Model with Multiscale Feature Fusion Capabilities for Open Water Aerial Imagery

Object detection algorithms for open water aerial images present challenges such as small object size, unsatisfactory detection accuracy, numerous network parameters, and enormous computational demands. Current detection algorithms struggle to meet the accuracy and speed requirements while being deployable on small mobile devices. This paper proposes DFLM-YOLO, a lightweight small-object detection network based on the YOLOv8 algorithm with multiscale feature fusion. Firstly, to solve the class imbalance problem of the SeaDroneSee dataset, we propose a data augmentation algorithm called Small Object Multiplication (SOM). SOM enhances dataset balance by increasing the number of objects in specific categories, thereby improving model accuracy and generalization capabilities. Secondly, we optimize the backbone network structure by implementing Depthwise Separable Convolution (DSConv) and the newly designed FasterBlock-CGLU-C2f (FC-C2f), which reduces the model’s parameters and inference time. Finally, we design the Lightweight Multiscale Feature Fusion Network (LMFN) to address the challenges of multiscale variations by gradually fusing the four feature layers extracted from the backbone network in three stages. In addition, LMFN incorporates the Dilated Re-param Block structure to increase the effective receptive field and improve the model’s classification ability and detection accuracy. The experimental results on the SeaDroneSee dataset indicate that DFLM-YOLO improves the mean average precision (mAP) by 12.4% compared to the original YOLOv8s, while reducing parameters by 67.2%. This achievement provides a new solution for Unmanned Aerial Vehicles (UAVs) to conduct object detection missions in open water efficiently.

Surface defect detection of strip steel based on GT-CutMix augmentation algorithm and improved DSSD model

Abstract Nowadays, defect detection technology based on deep learning continuously increases the surface quality requirements of hot-rolled strip steel. However, due to limitations in industrial production, defect datasets often suffer from insufficient training samples and imbalanced categories. This paper proposes effective solutions, namely the GT-CutMix offline data augmentation algorithm and lightweight small sample defect detection models. The GT-CutMix augmentation algorithm significantly improves defect utilization by accurately sampling defect locations and integrating them into the original data set. We design the S-DSSD defect detection model by constructing a lightweight SI-MobileNet to replace the ResNet101 backbone of the Deconvolutional Single Shot Detector (DSSD) network. This can reduce the resource parameters and consumption. At the same time, it can speed up training and inference. To further improve the detection accuracy, we integrate the Pyramid Split Attention (PSA) attention mechanism into the prediction module of DSSD and construct the SA-DSSD model. Under the GT-CutMix augmentation algorithm, the mAP of S-DSSD and SA-DSSD models on X-SDD dataset are 76.83% and 78.63%, respectively. Meanwhile, the corresponding detection speeds are 45 FPS and 40 FPS, respectively. In addition, on the NEU-DET cross-dataset experiment, the mAP of the SA-DSSD model reaches 74.88%. Our methods are highly effective and generalized for small sample defect detection, which can provide selective solutions for specific needs such as high speed and precision in different industrial production scenarios.

On circuit diameter bounds via circuit imbalances

We study the circuit diameter of polyhedra, introduced by Borgwardt, Finhold, and Hemmecke (SIAM J. Discrete Math. 29(1), 113–121 (2015)) as a relaxation of the combinatorial diameter. We show that the circuit diameter of a system {x∈Rn:Ax=b,0≤x≤u}\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\{x\\in \\mathbb {R}^n:\\, Ax=b,\\, \\mathbb {0}\\le x\\le u\\}$$\\end{document} for A∈Rm×n\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$A\\in \\mathbb {R}^{m\ imes n}$$\\end{document} is bounded by O(mmin{m,n-m}log(m+κA)+nlogn)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$O(m\\min \\{m, n - m\\}\\log (m+\\kappa _A)+n\\log n)$$\\end{document}, where κA\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\kappa _A$$\\end{document} is the circuit imbalance measure of the constraint matrix. This yields a strongly polynomial circuit diameter bound if e.g., all entries of A have polynomially bounded encoding length in n. Further, we present circuit augmentation algorithms for LPs using the minimum-ratio circuit cancelling rule. Even though the standard minimum-ratio circuit cancelling algorithm is not finite in general, our variant can solve an LP in O(mn2log(n+κA))\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$O(mn^2\\log (n+\\kappa _A))$$\\end{document} augmentation steps.

Alarm Log Data Augmentation Algorithm Based on a GAN Model and Apriori

Alarm Log Data Augmentation Algorithm Based on a GAN Model and Apriori

Multi-scenario surface temperature estimation in lithium-ion batteries with transfer learning and LGT augmentation

The State of Temperature (SOT) plays a crucial role in ensuring the safety and reliability of lithium-ion batteries, as well as the stability of electric vehicles (Evs). Recently, data-driven methods for lithium-ion battery temperature estimation have often obtained short-term estimation information in a single scenario. To alleviate the above issues, this paper proposes a method for estimating the surface temperature of lithium-ion batteries based on Local and Global Trend (LGT) data augmentation and Long Short-Term Memory (LSTM) model transfer. Initially, voltage and current data are processed using two types of filtering methods to reduce noise while preserving the fluctuating characteristics of the original data. Differential features are subsequently extracted from the polynomially filtered curves. Then all the data, including the filtered data and the features are fed into the LGT data augmentation algorithm, followed by the use of an LSTM transferable model to estimate the surface temperature of the target battery. Furthermore, the method is tested on both laboratory datasets and datasets with real driving conditions, covering a wide range of environmental temperatures and driving scenarios typical of EVs. The experimental results demonstrate that retraining with only the first 30 % of the target battery's data yields effective SOT estimation results.

Research on coal-rock identification method and data augmentation algorithm of comprehensive working face based on FL-Segformer

Coal-rock interface identification technology was pivotal in automatically adjusting the shearer’s cutting drum during coal mining. However, it also served as a technical bottleneck hindering the advancement of intelligent coal mining. This study aimed to address the poor accuracy of current coal-rock identification technology on comprehensive working faces, coupled with the limited availability of coal-rock datasets. The loss function of the SegFormer model was enhanced, the model’s hyperparameters and learning rate were adjusted, and an automatic recognition method was proposed for coal-rock interfaces based on FL-SegFormer. Additionally, an experimental platform was constructed to simulate the dusty environment during coal-rock cutting by the shearer, enabling the collection of coal-rock test image datasets. The morphology-based algorithms were employed to expand the coal-rock image datasets through image rotation, color dithering, and Gaussian noise injection so as to augment the diversity and applicability of the datasets. As a result, a coal-rock image dataset comprising 8424 samples was generated. The findings demonstrated that the FL-SegFormer model achieved a Mean Intersection over Union (MIoU) and mean pixel accuracy (MPA) of 97.72% and 98.83%, respectively. The FL-SegFormer model outperformed other models in terms of recognition accuracy, as evidenced by an MIoU exceeding 95.70% of the original image. Furthermore, the FL-SegFormer model using original coal-rock images was validated from No. 15205 working face of the Yulin test mine in northern Shaanxi. The calculated average error was only 1.77%, and the model operated at a rate of 46.96 frames per second, meeting the practical application and deployment requirements in underground settings. These results provided a theoretical foundation for achieving automatic and efficient mining with coal mining machines and the intelligent development of coal mines.

IRDA: Implicit data augmentation for deep imbalanced regression

Imbalanced data distributions are prevalent in real-world classification and regression tasks. Data augmentation is a commonly employed technique to mitigate this issue, with implicit methods gaining attention for their effectiveness and efficiency. However, implicit data augmentation methods have not been extensively explored in the context of regression tasks. To address this gap, we introduce IRDA, a novel learning method for regression that incorporates implicit data augmentation. Our approach includes developing a new augmentation strategy specifically tailored for deep imbalanced regression tasks, and a regression loss function that is suitable for real-world data with imbalanced label distributions and non-uniformly distributed features. We derive an easily computable surrogate loss and propose two implicit data augmentation algorithms, one incorporating meta-learning and one without. Additionally, we provide a regularization perspective to offer a deeper understanding of IRDA. We evaluate IRDA on five datasets, including a large-scale dataset, demonstrating its effectiveness in mitigating the adverse effects of imbalanced data distribution and its adaptability to various regression tasks.

Mitigating Large Language Model Bias: Automated Dataset Augmentation and Prejudice Quantification

Despite the growing capabilities of large language models, concerns exist about the biases they develop. In this paper, we propose a novel, automated mechanism for debiasing through specified dataset augmentation in the lens of bias producers that can be useful in a variety of industries, especially ones that are “restricted” and have limited data. We consider that bias can occur due to intrinsic model architecture and dataset quality. The two aspects are evaluated using two different metrics we created. We show that our dataset augmentation algorithm reduces bias as measured by our metrics. Our code can be found on an online GitHub repository.

Description and analysis of Sigfox received signal strength indicator dataset by using statistical techniques

Low power wide area network (LPWAN) technology has expanded and is essential in the development of applications for the internet of things (IoT). The Sigfox LPWAN network is characterized by its long-range coverage, low cost and power consumption. In this article, a set of 5174 values is analyzed, containing 1606 null RSSI data, obtained with the Sipy module and MicroPython, which provide a coverage map of several points with a resolution of 200 meters deployed in Quito–Ecuador. It is evaluated the type of distribution to which the set of network measurements is adjusted and an optimal 900 MHz propagation model in suburban environments is determined from the measurements obtained from the known base station. As a result, the lost values of RSSI were predicted using the inverse normal distribution method in the original values, observing that they conform to a logistic distribution. The data from the base station were subjected to a data augmentation algorithm designed in MATLAB, determining that the stanford university interim (SUI) model reduces the precision error in the trend of the curve by not presenting changes greater than 5 dB, achieving a precision of 97% with respect to the fit of the curve of the data.

A YOLO-based deep learning model for Real-Time face mask detection via drone surveillance in public spaces

Automating face mask detection in public areas is paramount to maintaining public health, especially in the context of the COVID-19 pandemic. Utilization of technologies such as deep learning and computer vision systems enables effective monitoring of mask compliance, thereby minimizing the risk of virus spread. Real-time detection helps in prompt intervention for and enforcement of the use of masks, thereby preventing potential outbreaks and ensuring compliance with public health guidelines. This method helps save human resources and makes the reinforcement of wearing masks in public areas consistent and objective. Automatic detection of face masks serves as a key tool for preventing the spread of contagious diseases, protecting public health, and creating a safer environment for every person. This study addresses the challenges of real-time face mask detection via drone surveillance in public spaces, with reference to three categories: wearing of mask, incorrect wearing of mask, and no mask. Addressing these challenges entails an efficient and robust object detection and recognition algorithm. This algorithm can deal with a crowd of multiple faces via a mobile camera carried by a mini drone, and performs real-time video processing. Accordingly, this study proposes a You Only Look Once (YOLO) based deep learning C-Mask model for real-time face mask detection and recognition via drone surveillance in public spaces. The C-Mask model aims to operate within a mini drone surveillance system and provide efficient and robust face mask detection. The C-Mask model performs preprocessing, feature extraction, feature generation, feature enhancement, feature selection, and multivariate classification tasks for each face mask detection cycle. The preprocessing task prepares the training and testing data in the form of images for further processing. The feature extraction task is performed using a Convolutional Neural Network (CNN). Moreover, Cross-Stage Partial (CSP) DarkNet53 is used to improve the feature extraction and to facilitate the model’s object detection ability. A data augmentation algorithm is used for feature generation to enhance the model’s training robustness. The feature enhancement task is performed by applying the Path Aggregation Network (PANet) and Spatial Pyramid Pooling Network (SPPNet) algorithms, which are deployed to enhance the extracted and generated features. The classification task is performed through multi-label classification, wherein each object in an image can belong to multiple classes simultaneously, and the network generates a grid of bounding boxes and corresponding confidence scores for each class. The YOLO-based C-Mask model testing is performed by experimenting with various face mask detection scenarios and with varying mask colors and types, to ensure the efficiency and robustness of the proposed model. The C-Mask model test results show that this model can correctly and effectively detect face masks in real-time video streams under various conditions with an overall accuracy of 92.20%, precision of 92.04, recall of 90.83%, and F1-score of 89.95%, for all the three classes. These high scores have been obtained despite mini drone mobility and camera orientation adjustment substantially affecting face mask detection performance.

Multi‐view synergistic enhanced fault recording data for transmission line fault classification

AbstractFault recorded data has been proven to be effective for fault diagnosis of overhead transmission lines. Utilizing deep learning to mine potential fault patterns in fault recording data is an inevitable trend. However, it is usually difficult to obtain massive labeled fault recording data, which results in deep learning‐based fault diagnosis models not being adequately trained. Although data augmentation methods provide ideas for expanding the training data, existing data augmentation algorithms (e.g. random perturbation‐based augmentation) may lead to distortion of multi‐view data, that is, time domain data and frequency domain data of the fault recorded data, which results in the inconsistency of physical properties and statistical distributions of the generated data and the actual recording data, and misguides the training of the models. Hence, this study proposes a transmission line fault classification method via the multi‐view synergistic enhancement of fault recording data. The methodology proposes to start with a synergistic enhancement of multi‐view data such as time and frequency domains of fault recording data, and utilizes contrastive learning to further improve the performance of the fault classification model while ensuring that the generated data is not distorted. Experimental results on three real‐world datasets validate the effectiveness of the proposed method.

Afitness: Fitness Monitoring on Smart Devices via Acoustic Motion Images

Recently, as fitness has become a popular part of people’s lives, the intention to record fitness processes and assess the standards of fitness movements has grown increasingly keen. However, the existing approaches have some limitations, for example, wearable devices can hinder users’ fitness activities; computer vision–based solutions pose the risk of privacy breach, and so on. Fortunately, we observed that smartspeaker, acoustic-based sensing is a promising method of activity monitoring. In this article, we propose Afitness, an acoustic-based sensing system that enables non-intrusive, passive, and high-precision fitness detection. Afitness has the following three innovations. (i) We utilize pulse compression to generate high-precision motion distance images on commercial devices that can be visually recognized. (ii) We propose a data augmentation algorithm, which also incorporates transfer learning to greatly reduce the pressure of data collection. (iii) We exploit incremental learning techniques that allow Afitness to improve the portability of our system and recognize new actions. Overall, Afitness achieves acoustic signal interpretability and environmental reliability detection.

OFIDA: Object-focused image data augmentation with attention-driven graph convolutional networks.

Image data augmentation plays a crucial role in data augmentation (DA) by increasing the quantity and diversity of labeled training data. However, existing methods have limitations. Notably, techniques like image manipulation, erasing, and mixing can distort images, compromising data quality. Accurate representation of objects without confusion is a challenge in methods like auto augment and feature augmentation. Preserving fine details and spatial relationships also proves difficult in certain techniques, as seen in deep generative models. To address these limitations, we propose OFIDA, an object-focused image data augmentation algorithm. OFIDA implements one-to-many enhancements that not only preserve essential target regions but also elevate the authenticity of simulating real-world settings and data distributions. Specifically, OFIDA utilizes a graph-based structure and object detection to streamline augmentation. Specifically, by leveraging graph properties like connectivity and hierarchy, it captures object essence and context for improved comprehension in real-world scenarios. Then, we introduce DynamicFocusNet, a novel object detection algorithm built on the graph framework. DynamicFocusNet merges dynamic graph convolutions and attention mechanisms to flexibly adjust receptive fields. Finally, the detected target images are extracted to facilitate one-to-many data augmentation. Experimental results validate the superiority of our OFIDA method over state-of-the-art methods across six benchmark datasets.

Adaptive Scale and Correlative Attention PointPillars: An Efficient Real-Time 3D Point Cloud Object Detection Algorithm

Recognizing 3D objects from point clouds is a crucial technology for autonomous vehicles. Nevertheless, LiDAR (Light Detection and Ranging) point clouds are generally sparse, and they provide limited contextual information, resulting in unsatisfactory recognition performance for distant or small objects. Consequently, this article proposes an object recognition algorithm named Adaptive Scale and Correlative Attention PointPillars (ASCA-PointPillars) to address this problem. Firstly, an innovative adaptive scale pillars (ASP) encoding method is proposed, which encodes point clouds using pillars of varying sizes. Secondly, ASCA-PointPillars introduces a feature enhancement mechanism called correlative point attention (CPA) to enhance the feature associations within each pillar. Additionally, a data augmentation algorithm called random sampling data augmentation (RS-Aug) is proposed to solve the class imbalance problem. The experimental results on the KITTI 3D object dataset demonstrate that the proposed ASCA-PointPillars algorithm significantly boosts the recognition performance and RS-Aug effectively enhances the training effects on an imbalanced dataset.