Search

The early detection and accurate rate of classification of plant leaf disease are more important for reliable and good agriculture, which prevents unwanted financial loss and resources. In this study, Tangent Hunter Prey Optimization-based LeNet (THPO-LeNet) is utilized for crop leaf classification and multi-class plant leaf disease identification. In this case, the input image that goes through the image pre-processing step is obtained from the plant village dataset. An adaptive Wiener filter is applied at the pre-processing stage to reduce needless mistakes and improve image quality. The pre-processed image is then sent to the leaf segmentation step, where a Mask Region-based Convolution Neural Network (Mask R-CNN) performs the segmentation. Consequently, image augmentation is performed using techniques such as scaling, rotation, translation, flipping, contrast, saturation, and hue. Afterwards, first-level classification plant leaf classification is processed by LeNet, which is optimized by Tangent Hunter Prey Optimization (THPO). The THPO is the incorporation of a Tangent Search Algorithm (TSA) and Hunter–Prey Optimizer (HPO). At last, the second-level classification of plant leaf disease is conducted by THPO-LeNet. Furthermore, the efficiency of THPO-LeNet is examined based on measures, like accuracy, Negative Predictive Value (NPV), True Positive Rate (TPR), True Negative Rate (TNR), Positive Predictive Value (PPV), loss value, and False Positive Rate (FPR), and the value attained is 95.68%, 92.50%, 91.48%, 92.25%, 92.54%, 8.321%, and 7.754%, respectively. [ABSTRACT FROM AUTHOR]

Objectives Recently, deep learning medical image analysis in orthopedics has become highly active. However, progress has been restricted by the absence of large-scale and standardized ground-truth images. To the best of our knowledge, this study is the first to propose an innovative solution, namely a deep few-shot image augmentation pipeline, that addresses this challenge by synthetically generating knee radiographs for training downstream tasks, with a specific focus on knee osteoarthritis Kellgren-Lawrence (KL) grading. Materials and Methods This study leverages a deep few-shot image augmentation pipeline to generate synthetic knee radiographs. Despite the limited availability of training samples, we demonstrate the capability of our proposed computational strategy to produce high-fidelity plain knee radiographs and use them to successfully train a KL grade classifier. Results Our experimental results showcase the effectiveness of the proposed computational pipeline. The generated synthetic radiographs exhibit remarkable fidelity, evidenced by the achieved average Frechet Inception Distance (FID) score of 26.33 for KL grading and 22.538 for bilateral knee radiographs. For KL grading classification, the classifier achieved a test Cohen's Kappa and accuracy of 0.451 and 0.727, respectively. Our computational strategy also resulted in a publicly and freely available imaging dataset of 86 000 synthetic knee radiographs. Conclusions Our approach demonstrates the capability to produce top-notch synthetic knee radiographs and use them for KL grading classification, even when working with a constrained training dataset. The results obtained emphasize the effectiveness of the pipeline in augmenting datasets for knee osteoarthritis research, opening doors for broader applications in orthopedics, medical image analysis, and AI-powered diagnosis. [ABSTRACT FROM AUTHOR]

Soybean diseases pose a significant obstacle to the increase in worldwide demand for soybean, leading to significant reductions in production. This study introduces an innovative approach that utilizes convolution neural network (CNN) and the internet of things (IoT) to efficiently identify and categorize various types of soybean leaf infections. A deep CNN model has been developed using the AlexNet architecture to effectively classify soybean plant diseases. A collection of 11,493 leaf images of 14 soybean diseases, including healthy plants was acquired from the ground using IoT devices and camera modules. Achieving 99.39% and 97.01% accuracy and F1-score respectively, helps not only in early disease detection, but also brings about a transformative approach to sustainable soybean crop management. This initiative lays the foundation for a strengthened agricultural future, enabling increased crop production and improved economic well-being in the cultivation of soybean, all facilitated by the integration of innovative technologies. [ABSTRACT FROM AUTHOR]

Wildfires pose significant environmental and societal threats, necessitating improved early detection methods. This study investigates the effectiveness of integrating real-time object detection deep learning models (YOLOv8 and RT-DETR) with advanced data augmentation techniques, including StyleGAN2-ADA, for wildfire smoke detection. We evaluated model performance on datasets enhanced with fundamental transformations and synthetic images, focusing on detection accuracy. YOLOv8X demonstrated superior overall performance with AP@0.33 of 0.962 and AP@0.5 of 0.900, while RT-DETR-X excelled in small object detection with a 0.983 detection rate. Data augmentation, particularly StyleGAN2-ADA, significantly enhanced model performance across various metrics. Our approach reduced average detection times to 1.52 min for YOLOv8X and 2.40 min for RT-DETR-X, outperforming previous methods. The models demonstrated robust performance under challenging conditions, like fog and camera noise, providing reassurance of their effectiveness. While false positives remain a challenge, these advancements contribute significantly to early wildfire smoke detection capabilities, potentially mitigating wildfire impacts through faster response times. This research establishes a foundation for more effective wildfire management strategies and underscores the potential of deep learning applications in environmental monitoring. [ABSTRACT FROM AUTHOR]

BACKGROUND: With the rapid development of deep learning, the recognition of rice disease images using deep neural networks has become a hot research topic. However, most previous studies only focus on the modification of deep learning models, while lacking research to systematically and scientifically explore the impact of different data sizes on the image recognition task for rice diseases. In this study, a functional model was developed to predict the relationship between the size of dataset and the accuracy rate of model recognition. RESULTS: Training VGG16 deep learning models with different quantities of images of rice blast‐diseased leaves and healthy rice leaves, it was found that the test accuracy of the resulting models could be well fitted with an exponential model (A = 0.9965 − e(−0.0603×I50−1.6693)). Experimental results showed that with an increase of image quantity, the recognition accuracy of deep learning models would show a rapid increase at first. Yet when the image quantity increases beyond a certain threshold, the accuracy of image classification would not improve much, and the marginal benefit would be reduced. This trend remained similar when the composition of the dataset was changed, no matter whether (i) the disease class was changed, (ii) the number of classes was increased or (iii) the image data were augmented. CONCLUSIONS: This study provided a scientific basis for the impact of data size on the accuracy of rice disease image recognition, and may also serve as a reference for researchers for database construction. © 2024 Society of Chemical Industry. [ABSTRACT FROM AUTHOR]

Two key building characteristics, namely the number of stories and typology, is vital across various domains such as construction management and architectural design. These aspects are particularly critical for disaster risk assessment and infrastructure planning. Although deep learning models are adept at extracting this information from Street view images (SVIs), their success is contingent upon the availability of large and diverse data sets with high accuracy. Image augmentation presents an alternative method to artificially broaden data set diversity. However, the impact of image augmentation techniques on identifying building stories and typologies from SVIs has not been adequately explored. This study proposes a methodology employing eight distinct image augmentation techniques—brightness, contrast, perspective, rotation, scale, shearing, and translation augmentations—as well as a combined approach using all these methods. The study evaluates the efficacy of models trained with these techniques by comparing the accuracy of different classes and architectures for each task, both with and without the application of augmentation. The findings revealed that while most augmentation methods enhance model accuracy, their effectiveness is task-dependent. Furthermore, it was observed that the most effective augmentation techniques differ among building classes and architectures within each task. This suggests that augmentation strategies need to be custom-designed to align with the unique features of each class and architectures for precise estimation of the number of stories and building typologies. While the focus of this research is on specific tasks, the evaluated augmentation techniques could also extend to related areas, such as ascertaining the age of buildings or identifying window types. Practical Applications: In this study, the efficacy of augmentation techniques is explored within the framework of identifying the number of stories and building typologies. The models were assessed for average accuracy and class-specific accuracy across various architectures, comparing outcomes with and without the implementation of the proposed augmentation methods. A key finding is that the most effective augmentation method varies between architectures and individual classes. Contrary to common practice in deep learning, where applying multiple augmentation techniques is standard for accuracy enhancement, this study observed that such a strategy did not uniformly improve performance. Specifically, while combining augmentation methods generally resulted in higher average accuracy, this was not the case for some classes within MobileNetV3 when detecting the number of stories. Similarly, for ResNet-152, employing all augmentation techniques together led to the lowest accuracy in certain classes for building typology classification. These results indicate that augmentation strategies may require customization to cater to the distinct characteristics of each class and architecture for accurate estimation of number of stories and building typologies. [ABSTRACT FROM AUTHOR]

Fashion items are hard to classify since there are a million variations in style, texture, and pattern. Image classification is among the noted strengths of convolutional neural networks. This research introduces an improved CNN architecture for fashion classification, utilizing image augmentation and batch normalization to improve model performance and generalization. To make the model more robust, image augmentation techniques like rotation, width and height shift, zoom, and flips were employed. In addition, a Batch Normalization layer is added in the middle, which can help on stabilizing the learning process and accelerating convergence. The proposed model was trained on an augmented dataset, achieving a satisfactory improvement in test accuracy of 91.97% compared to a baseline CNN model, which obtained 88.5% accuracy. According to the results, the image augmentation with the application of Batch Normalization improves the CNN architecture for better effectiveness in fashion classification tasks. [ABSTRACT FROM AUTHOR]

This study explores the utilization of deep learning techniques for accurate cashew classification to improve efficiency and accuracy in the cashew industry. YOLOv5, YOLOv9, and a Convolutional Neural Network (CNN) were evaluated in classifying cashews into whole, broken, split-up, split-down, and defect categories. A comprehensive labeled dataset was built to train the models, applying data augmentation to increase robustness. YOLOv5 achieved the highest accuracy of 97.65% and the fastest inference time (0.025 s per image) among the models, making it suitable for real-time applications. Although CNN offered a simpler architecture, YOLOv5's superior performance places it as the most promising candidate for large-scale cashew classification deployment. [ABSTRACT FROM AUTHOR]

Agriculture is India’s most common job, yet it lacks innovation and technology. As the world’s population expands, so does the demand for more food. Pesticides are used on farms to boost yield. The toxicity of the fertilizer has serious health repercussions for the farmer. So, it’s recommended to measure the amount of pesticide used and only apply it when necessary. We devised an insect-finding and insecticide-spraying mechanism. This is accomplished by employing a drone or Uninterrupted Ariel Vehicle. The drone has a camera that can photograph fields and lift pesticides weighing 3 to 4kg. After locating the insect, the insecticide is sprayed through the nozzles. In the proposed model, the Deep Convolutional Neural Network (CNN) has reached state of the art in image processing and object detection issues. Deep CNN has the potential to self-learn hidden features that help with insect detection. When compared to other similar approaches, experimental findings on a real dataset to illustrate the usefulness of the suggested methodology. We identified insects on the crop with 90% accuracy using deep CNN. It helps farmers to increase crop yield while also shielding them from the detrimental effects of spraying pesticides on the field manually. [ABSTRACT FROM AUTHOR]

With the rise of big data processing is creating a new challenge i.e., keeping pace with the flow of information and efficient I/O performance is the key here. However, analyzing I/O performance is a complex task due to the many layers involved, from applications and libraries to the operating system, storage devices, and everything in between. In this paper, we propose a convolutional neural networks (CNN)-based directed acyclic graph regression (DAGR) network to predict the I/O performance of applications. The system first gathers I/O request information directly from the storage layer (block storage). This information is then converted into a visual representation (graph image) and augmented using various techniques to create additional training data. The core of the system is a CNN-based prediction model designed to identify potential I/O performance patterns by analyzing the generated graph images. Evaluations using real-world application benchmarks demonstrate that the proposed method can accurately predict the performance of various applications, including file servers, databases, mail servers, and video servers, with an accuracy of up to 99.73%. [ABSTRACT FROM AUTHOR]

In Human-Computer Interaction, lip reading is essential and still an open research problem. In the last decades, there have been many studies in the field of Automatic Lip-Reading (ALR) in different languages, which is important for societies where the essential applications developed. Similarly to other machine learning and artificial intelligence applications, Deep Learning (DL) based classification algorithms have been applied for ALR in order to improve the performance of ALR. In the field of ALR, few studies have been done on the Turkish language. In this study, we undertook a multifaceted approach to address the challenges inherent to Turkish lip reading research. To begin, we established a foundation by creating an original dataset meticulously curated for the purpose of this investigation. Recognizing the significance of data quality and diversity, we implemented three robust image data augmentation techniques: sigmoidal transform, horizontal flip, and inverse transform. These augmentation methods not only elevated the quality of our dataset but also introduced a rich spectrum of variations, thereby bolstering the dataset's utility. Building upon this augmented dataset, we delved into the application of cuttingedge DL models. Our choice of models encompassed Convolutional Neural Networks (CNN), known for their prowess in extracting intricate visual features, Long-Short Term Memory (LSTM), adept at capturing sequential dependencies, and Bidirectional Gated Recurrent Unit (BGRU), renowned for their effectiveness in handling complex temporal data. These advanced models were selected to leverage the potential of the visual Turkish lip reading dataset, ensuring that our research stands at the forefront of this rapidly evolving field. The dataset utilized in this study was gathered with the primary objective of augmenting the extant corpus of Turkish language datasets, thereby substantively enriching the landscape of Turkish language research while concurrently serving as a benchmark reference. The performance of the applied method has been compared regarding precision, recall, and F1 metrics. According to experiment results, BGRU and LSTM models gave the same results up to the fifth decimal, and BGRU had the fastest training time. [ABSTRACT FROM AUTHOR]

[Objectives] Identifying crops and weeds are crucial aspects of advancing intelligent and automated weeding. This article aimed to improve the accuracy of crop and weed identification, to enhance the real-time performance of detection models, and to enhance robustness. [Methods] Focusing on maize crops and their corresponding weeds in the leaf age range of 3-8 leaves, this research endeavored to devise a detection method for maize seedlings and associated grasses. The seedling detection method leveraged an improved real-time end-to-end object detection with transformers (RT-DETR) for maize and weed detection in field conditions. The novel concept of replacing large-scale deep convolution with small-scale convolution equivalence within RT-DETR was introduced, reducing training complexity and inference time while maintaining detection accuracy. Furthermore, a self-attention mechanism with contextual information was integrated to enhance target attention and improve small target detection. Additionally, a combined image enhancement strategy was employed to enhance model accuracy and generalization. [Results]The improved model effectively distinguished weeds from crops in complex field scenarios, achieving an average detection accuracy of 90.11%. In the inference stage, each image took 33.67 ms for processing, with a model size of 44.86 MB. Compared with the mainstream target detection model, the improved model had higher overall accuracy and fast speed. [Conclusions]The proposed method had excellent overall detection performance for corn seedlings and associated weeds, which could improve the accuracy and efficiency of weed identification. [ABSTRACT FROM AUTHOR]

Medical imaging is essential to modern life support. Different approaches and standards exist for detecting life-threatening disorders. However, manually detecting anomalies takes time and is prone to error. Additionally, it demands great expertise and competence. Deep learning in medical imaging improves accuracy and provides several benefits. Privacy concerns limit access to medical data, especially images. This problem can be solved with data augmentation. Image quality determines prediction accuracy. Preprocessing images improves quality. This article addresses data augmentation and image enhancement with a multiscale self-attention generator. Foundational network is Inception. Elastic_net used for selection and regularisation of features. The anisotropic diffusion filter used for reducing speckle noise in the model. The ZONODO foetal planes dataset is used as input. The proposed model produces better results for classification as accuracy is 98.7, precision is 97.5, recall is 97.8 and F1-Score is 97.5 when compared with existing methods. [ABSTRACT FROM AUTHOR]

In this study, we developed and explored a methodical image augmentation technique for swimmer localisation in northern German outdoor lake environments. When it comes to enhancing swimmer safety, a main issue we have to deal with is the lack of real-world training data of such outdoor environments. Natural lighting changes, dynamic water textures, and barely visible swimming persons are key issues to address. We account for these difficulties by adopting an effective background removal technique with available training data. This allows us to edit swimmers into natural environment backgrounds for use in subsequent image augmentation. We created 17 training datasets with real images, synthetic images, and a mixture of both to investigate different aspects and characteristics of the proposed approach. The datasets were used to train YOLO architectures for possible future applications in real-time detection. The trained frameworks were then tested and evaluated on outdoor environment imagery acquired using a safety drone to investigate and confirm their usefulness for outdoor swimmer localisation. [ABSTRACT FROM AUTHOR]

In this study, we developed and explored a methodical image augmentation technique for swimmer localisation in northern German outdoor lake environments. When it comes to enhancing swimmer safety, a main issue we have to deal with is the lack of real-world training data of such outdoor environments. Natural lighting changes, dynamic water textures, and barely visible swimming persons are key issues to address. We account for these difficulties by adopting an effective background removal technique with available training data. This allows us to edit swimmers into natural environment backgrounds for use in subsequent image augmentation. We created 17 training datasets with real images, synthetic images, and a mixture of both to investigate different aspects and characteristics of the proposed approach. The datasets were used to train YOLO architectures for possible future applications in real-time detection. The trained frameworks were then tested and evaluated on outdoor environment imagery acquired using a safety drone to investigate and confirm their usefulness for outdoor swimmer localisation.

Despite the medicinal significance of traditional Thai herbs, their accurate identification on digital platforms is a challenge due to the vast diversity among species and the limited scope of existing digital databases. In response, this paper introduces the Thai traditional herb classifier that uniquely combines transfer learning, innovative data augmentation strategies, and the inclusion of noisy data to tackle this issue. Our novel contributions encompass the creation of a curated dataset spanning 20 distinct Thai herb categories, a robust deep learning architecture that intricately combines transfer learning with tailored data augmentation techniques, and the development of an Android application tailored for real-world herb recognition scenarios. Preliminary results of our method indicate its potential to revolutionize the way Thai herbs are digitally identified, holding promise for advancements in natural medicine and computer-assisted herb recognition. [ABSTRACT FROM AUTHOR]

Highlights: The artificial neural network-based computer vision technique can be used to automatically classify rock joint profiles with reasonably good classification accuracies. A total of 200 images of rock joints are used to train and test the proposed method. This method overcomes drawbacks of the conventional visual assessment, leading to reduced subjectivity and costs, and rapid estimation of rock joint profile roughness. [ABSTRACT FROM AUTHOR]

We propose an effective strategy for detecting and localizing anomalous behavior using a modified end-to-end two-stage encoder-decoder U-shaped network. By building the model from scratch for the detection, segmentation, and classification of an anomalous event in video sequences. The encoder model is helpful for feature extraction which is input to the bottleneck block. The resulting feature maps serve as input to the decoder path which is responsible for transposing feature maps to the original image. For precise localization of anomaly, we included the augmentation feature of both images and their masks in the proposed U-net model. In a two-stage U-net network, the first model is useful for the detection of video frames while the same U-net model is used in the second stage for augmentation of detected video frames from first model, which provides segmentation and classification of images. This precise symmetric path-based architecture is useful in good spatial localization of anomalous events. We apply a pixel-based threshold value for Intersection over Union score to distinguish the pixels. Pixels having values greater than the threshold are considered anomalous or otherwise normal with an IoU score of 0. We have evaluated our two-stage U-net model performance on three benchmark standard datasets and compared performance with the Conventional U-net model, and Attention-U-net models without augmentation features. Our method combines spatial details and deep features that yield an improved accuracy of 99.15%, a mean intersection over union score of 82.33 and 99% ROC values that are higher as compared to other methods. [ABSTRACT FROM AUTHOR]

In binder jetting followed by sintering, the porosity characterization is critical to understand how the process affects the structure of the printed parts. Image-based porosity detection methods are widely used but the current solutions are limited to specific materials and conditions and require manual tuning that precludes real-time porosity detection. The application of machine learning for automating porosity detection has been also limited to specific materials and conditions and requires a large training dataset for successful implementation. However, large datasets are difficult to acquire experimentally in binder jetting due to prohibited material costs and experiment time. To bridge the knowledge gap, this paper investigates the application of machine learning on automated porosity detection using a small dataset consisting of highly varied cross-section images of metal parts produced by binder jetting followed by sintering. Stylegan3, a type of generative adversarial network, is used to increase the number of training images by image augmentation, and YOLOv5, a convolutional neural network specialized for object detection, is used to detect porosities. The resulting model achieves an F1 score of 88% and detection time of 3–15 ms per image. Generalized porosity detection is also assessed on a set of images containing highly varied materials, resolutions, magnifications, and pore densities. Furthermore, morphological information of the classified porosities such as the distribution of their orientations are automatically extracted using image processing algorithms. [ABSTRACT FROM AUTHOR]

Deep Learning models are widely used to address COVID-19 challenges, but they require a large number of training samples. X-ray images of COVID-19 patients are amongst the preferred methods for detection. However, their availability is limited. In contrast, X-ray images of non-COVID-19 patients are available in abundance. Furthermore, COVID-19 patient's treatment varies based on infection severity. This leads to a class imbalance issue as there are far more X-ray images of non-COVID patients than COVID-19 patients available for training deep learning models. As a result, deep learning models cannot achieve the desired levels of accuracy. This study's primary objective is to generate synthetic X-ray images depicting three levels of severity, utilizing a Cycle Consistent Generative Adversarial Network (CycleGAN). The Structural Similarity Index (SSIM) is employed to create training datasets for three severity levels, which are then used to train the corresponding CycleGAN models. Additionally, a comparative analysis is conducted to compare the achieved accuracies between X-ray images of COVID-19 patients and non-COVID-19 patients. This analysis involves datasets containing authentic non-synthetic COVID-19 X-ray images and datasets containing synthetic COVID-19 X-ray images generated using CycleGAN. The results indicate enhanced accuracy when deep models are trained using augmented X-ray data. This study is novel as no prior work has been done on severity-wise dataset generation and classification of COVID-19 X-ray images. [ABSTRACT FROM AUTHOR]

Accurate inspection of rebars in Reinforced Concrete (RC) structures is essential and requires careful counting. Deep learning algorithms utilizing object detection can facilitate this process through Unmanned Aerial Vehicle (UAV) imagery. However, their effectiveness depends on the availability of large, diverse, and well-labelled datasets. This article details the creation of a dataset specifically for counting rebars using deep learning-based object detection methods. The dataset comprises 874 raw images, divided into three subsets: 524 images for training (60 %), 175 for validation (20 %), and 175 for testing (20 %). To enhance the training data, we applied eight augmentation techniques—brightness, contrast, perspective, rotation, scale, shearing, translation, and blurring—exclusively to the training subset. This resulted in nine distinct datasets: one for each augmentation technique and one combining all techniques in augmentation sets. Expert annotators labelled the dataset in VOC XML format. While this research focuses on rebar counting, the raw dataset can be adapted for other tasks, such as estimating rebar diameter or classifying rebar shapes, by providing the necessary annotations.

This data article presents a comprehensive dataset comprising breast cancer images collected from patients, encompassing two distinct sets: one from individuals diagnosed with breast cancer and another from those without the condition. Expert physicians carefully select, verify, and categorize the dataset to guarantee its quality and dependability for use in research and teaching. The dataset, which originates from Sulaymaniyah, Iraq, provides a distinctive viewpoint on the frequency and features of breast cancer in the area. This dataset offers a wealth of information for developing and testing deep learning algorithms for identifying breast cancer, with 745 original images and 9,685 augmented images. The addition of augmented X-rays to the dataset increases its adaptability for algorithm development and instructional projects. This dataset holds immense potential for advancing medical research, aiding in the development of innovative diagnostic tools, and fostering educational opportunities for medical students interested in breast cancer detection and diagnosis.

Visual defect classification is a critical process in manufacturing systems, aiming to achieve high-quality production and reduce costs. Although deep learning-based defect classification models have achieved significant success, their performance can be significantly diminished due to ‘environment shifts’—variations in manufacturing environments across multiple production lines. To address this challenge, we propose a domain augmentation framework to construct an environmentally-robust defect classification model, delivering high performance across various manufacturing environments using a training dataset from only a single production line. In our framework, each environment is treated as a separate domain, and multiple augmented domains are created using image transformation functions. Subsequently, a defect classification model is trained using a multi-source domain generalization (DG) method with these augmented domains. This approach mitigates the single-source DG problem to a multi-source DG problem, enabling the adoption of multi-source DG methods, which leads to performance improvements. The effectiveness of the proposed framework is demonstrated through experiments on a dataset provided by a Korean manufacturing company.

Goal: In recent years, deep neural networks have consistently outperformed previously proposed methods in the domain of medical segmentation. However, due to their nature, these networks often struggle to delineate desired structures in data that fall outside their training distribution. The goal of this study is to address the challenges associated with domain generalization in CT segmentation by introducing a novel method called BucketAugment for deep neural networks. Methods: BucketAugment leverages principles from the Q-learning algorithm and employs validation loss to search for an optimal policy within a search space comprised of distributed stacks of 3D volumetric augmentations, termed ‘buckets.’ These buckets have tunable parameters and can be seamlessly integrated into existing neural network architectures, offering flexibility for customization. Results: In our experiments, we focus on segmenting kidney and liver structures across three distinct medical datasets, each containing CT scans of the abdominal region collected from various clinical institutions and scanner vendors. Our results indicate that BucketAugment significantly enhances domain generalization across diverse medical datasets, requiring only minimal modifications to existing network architectures. Conclusions: The introduction of BucketAugment provides a promising solution to the challenges of domain generalization in CT segmentation. By leveraging Q-learning principles and distributed stacks of 3D augmentations, this method improves the performance of deep neural networks on medical segmentation tasks, demonstrating its potential to enhance the applicability of such models across different datasets and clinical scenarios.

One of the most prevalent forms of malignant tumor affecting the digestive system is the Colorectal Cancer (CRC). Due to the recurrent nature of occurrence of CRC tumors the morbidity and mortality rate associated with CRC tumors is very high, making it the fourth leading cause of cancer worldwide. Although colonoscopy, is considered as pre-dominant screening mechanism for detection of CRC, over the recent years several researchers have attempted in developing AI diagnostic tools for segmenting colonic polyps. However, manual intervention and decreased accuracy rate is a major drawback witnessed by these approaches. In this study, we propose a novel automated colonic polyp segmentation mechanism using U-Net and chroma based deformable model termed as ContourNet. The proposed model eliminates the need for manual identification of region of interest irrespective of the high degree of variability between the lesion and non-lesion pixels. The model considers the chromaticity and statistical information from the identified region of interest to evolve the contour close to the affected area. The proposed algorithm was developed using Clinic CVC dataset for training, and a test accuracy of 94% was obtained. The generalization ability of the proposed design is validated on three different datasets, considering all the images as test set, a high degree of accuracy was obtained in identifying the affected regions. Thereby the results obtained prove the ability of the proposed method to be adopted in developing a CAD tool in detection of pathologies associated with gastroenterology in contrast to the state of art methods reported in literature.

Wildfires pose significant environmental and societal threats, necessitating improved early detection methods. This study investigates the effectiveness of integrating real-time object detection deep learning models (YOLOv8 and RT-DETR) with advanced data augmentation techniques, including StyleGAN2-ADA, for wildfire smoke detection. We evaluated model performance on datasets enhanced with fundamental transformations and synthetic images, focusing on detection accuracy. YOLOv8X demonstrated superior overall performance with AP@0.33 of 0.962 and AP@0.5 of 0.900, while RT-DETR-X excelled in small object detection with a 0.983 detection rate. Data augmentation, particularly StyleGAN2-ADA, significantly enhanced model performance across various metrics. Our approach reduced average detection times to 1.52 min for YOLOv8X and 2.40 min for RT-DETR-X, outperforming previous methods. The models demonstrated robust performance under challenging conditions, like fog and camera noise, providing reassurance of their effectiveness. While false positives remain a challenge, these advancements contribute significantly to early wildfire smoke detection capabilities, potentially mitigating wildfire impacts through faster response times. This research establishes a foundation for more effective wildfire management strategies and underscores the potential of deep learning applications in environmental monitoring.

Agriculture is one of the most crucial aspects of a nation's growth. However, the quality and quantity of crop yield are severely affected by various plant diseases. Plant diseases must be identified and prevented at an early stage to improve food quality and production rate. The emerging deep learning network of convolutional neural networks (CNNs) achieved excellent results in plant disease classification. However, the classification potential of the network depends largely on the configuration of hyperparameters. Finding the optimal set of hyperparameters is a tedious, time-consuming, and challenging task. To tackle such an issue, this paper proposes an optimized CNN integrated with a novel modified whale optimization algorithm (MWOA) to achieve plant disease classification. Here, the novel method of optimizing CNN hyperparameters facilitates quicker implementation. To boost the proposed model's efficiency, several data augmentation methods are used such as rotation, scaling, and flipping. The proposed MWOA-CNN model is implemented using a plant village database that includes fourteen plant species, 38 disease classes, and healthy leaves as well. The experiential findings showed that the proposed model outperforms the existing models by attaining the highest classification accuracy of 99.92%, resulting in an effective model for plant disease classification. [ABSTRACT FROM AUTHOR]

This paper introduces a noise augmentation technique designed to enhance the robustness of state-of-the-art (SOTA) deep learning models against degraded image quality, a common challenge in long-term recording systems. Our method, demonstrated through the classification of digital holographic images, utilizes a novel approach to synthesize and apply random colored noise, addressing the typically encountered correlated noise patterns in such images. Empirical results show that our technique not only maintains classification accuracy in high-quality images but also significantly improves it when given noisy inputs without increasing the training time. This advancement demonstrates the potential of our approach for augmenting data for deep learning models to perform effectively in production under varied and suboptimal conditions. [ABSTRACT FROM AUTHOR]

Image manipulation has become a common problem due to the tremendous growth of digital tools and applications. Several image forgeries can uneventfully eradicate the original contents from digital images. Copy-paste forgeries become the most challenging problem, where the image's content is manipulated by copying and pasting from one region to another location within the same image. Existing forgery detection techniques face a high complexity in identifying the manipulated region when it is subjected to various geometrical alterations. Hence, this study introduces a novel deep learning (DL) based technique to detect the copy paste forgeries in digital images. This research undergoes three major operations pre-processing, image augmentation and classification. Image normalization, rescaling, and error level analysis (ELA) are performed in the pre-processing phase, which can improve accuracy performance. In addition, the proposed pre-processing technique can reduce the overfitting issues in the network model. Then, the image augmentation is conquered to maximize the size of the dataset images. Finally, the convolutional Autoencoder based deep learning (DL) technique is proposed to accurately classify the fake image. The MICC-F220 dataset is utilized for experimentation, and the proposed method is processed using the PYTHON platform. The proposed method achieves an overall accuracy of 99.2%, a specificity of 96.5%, a recall of 95.79% and an F1 score of 96.09%. In addition, the proposed method is compared with conventional techniques and proves the system's efficiency. [ABSTRACT FROM AUTHOR]

Construction materials undergo appearance and textural changes during the construction process. Accurate recognition of these changes is critical for effectively understanding the construction status; however, recognizing the various levels of detailed material conditions is not sufficiently explored. The primary challenge in the detailed recognition of the conditions of the material is the availability of labeled training data. To address this challenge, this study proposes a novel state-of-the-art deep learning model that leverages transfer learning, utilizing the pretrained Inception V3 to transfer knowledge to the limited labeled data set in the construction context. This enables the model to learn meaningful representations from the limited training data, enhancing its ability to accurately classify material conditions. In addition, gray-level co-occurrence matrix (GLCM)–based texture features are extracted from the images to capture the appearance and textural changes in construction materials, which are then concatenated with the transferred convolutional neural network (CNN) features to create a more comprehensive representation of the material conditions. The proposed model achieved an overall classification accuracy of 95% and 71% with limited (208 images) and very small (70 images) data sets, respectively. It outperformed different experimental architectures, including CNN models developed using limited data with and without augmentation, CNN model with data augmentation and transfer learning, separate models using local binary pattern (LBP) and GLCM texture features with super learners trained using augmented limited data. The findings suggest that the proposed model, which combines transfer learning with GLCM-based texture features, is effective in accurately recognizing the conditions of construction materials, even with limited labeled training data. This can contribute to improved construction management and monitoring. There are several practical applications of the proposed model combining CNN architecture with GLCM textures in construction industry. The primary applications are in activities such as quality inspection and progress monitoring. By analyzing images of different material conditions, such as concrete, plaster, or masonry walls, the model can be used to automatically detect defects or inconsistencies. This enables construction practitioners to ensure the quality of their structures, detect issues early on, and make informed decisions for maintenance and repair. Additionally, the model can be used to monitor the progress of construction projects by analyzing images to track the status and completion of different building components to estimate delays or cost overruns comparing with the expected material condition at a given time of the schedule. Moreover, because the model does not depend on large training data set, it enables construction managers to develop their project specific data sets and automate material condition detection and monitoring. [ABSTRACT FROM AUTHOR]

In the manufacturing industry, which is facing the 4th Industrial Revolution, various process data are being collected from various sensors, and efforts are being made to construct more efficient processes using these data. Many studies have demonstrated high accuracy in predicting defect rates through image data collected during the process using two-dimensional (2D) convolutional neural network (CNN) algorithms, which are effective in image analysis. However, in an environment where numerous process data are recorded as numerical values, the application of 2D CNN algorithms is limited. Thus, to perform defect prediction through the application of a 2D CNN algorithm in a process wherein image data cannot be collected, this study attempted to develop a defect prediction technique that can visualize the data collected in numerical form. The polyurethane foam manufacturing process was selected as a case study to verify the proposed method, which confirmed that the defect rate could be predicted with an average accuracy of 97.32%. Consequently, highly accurate defect rate prediction and verification of the basis of judgment can be facilitated in environments wherein image data cannot be collected, rendering the proposed technique applicable to processes other than those in this case study. [ABSTRACT FROM AUTHOR]

Generative Adversarial Networks (GANs) have been widely applied in various domains, including medical image analysis. GANs have been utilized in classification and segmentation tasks, aiding in the detection and diagnosis of diseases and disorders. However, medical image datasets often suffer from insufficiency and imbalanced class distributions. To overcome these limitations, researchers have employed GANs to generate augmented medical images, effectively expanding datasets and balancing class distributions. This review follows the PRISMA guidelines and systematically collects peer-reviewed articles on the development of GAN-based augmentation models. Automated searches were conducted on electronic databases such as IEEE, Scopus, Science Direct, and PubMed, along with forward and backward snowballing. Out of numerous articles, 52 relevant ones published between 2018 and February 2022 were identified. The gathered information was synthesized to determine common GAN architectures, medical image modalities, body organs of interest, augmentation tasks, and evaluation metrics employed to assess model performance. Results indicated that cGAN and DCGAN were the most popular GAN architectures in the reviewed studies. Medical image modalities such as MRI, CT, X-ray, and ultrasound, along with body organs like the brain, chest, breast, and lung, were frequently used. Furthermore, the developed models were evaluated, and potential challenges and future directions for GAN-based medical image augmentation were discussed. This review presents a comprehensive overview of the current state-of-the-art in GAN-based medical image augmentation and emphasizes the potential advantages and challenges associated with GAN utilization in this domain. [ABSTRACT FROM AUTHOR]

There is a need for robust methodological approaches to improve our capacity to automatically detect plant species from seed samples tohelp support plant management strategies. In this study, we tested different neural network techniques to automatically detect native species from seeds from the Andean páramo region based on optic RGB images. Specifically, we compared i) simple feed-forward networks (SNNs), consisting of feed-forward nets with error back-propagation, holding one hidden layer with different number of neurons; and ii) deep convolutional neural networks (CNNs), which have their convolutional layers-built form multiple 3x3 kernels. First, we sampled 50 seeds from four common plant species in the La Rusia Páramo (Colombia): Espeletia congestiflora, Bucquetia glutinosa, Calamagrostis effusa and Puya santosii. We took RGB images of individual seeds for each species on contrasted white and black backgrounds, and then classified all images under both SNNs and CNNs. Under a double cross-validation scheme, the SNN approach with 14 neurons approached 88% of test accuracy, while CNN achieved 93%. Moreover, when increasing the image sample in the training dataset fed to models, CNN performed with 100% accuracy when used on testing and validation datasets. Overall, the neural network approach explored here suggests a promising methodology for species prediction from seeds based on optical RGB images, with potential for automatic seed recognition and counting on the field. [ABSTRACT FROM AUTHOR]

Hand gesture recognition (HGR) is a challenging and fascinating research topic in computer vision with numerous daily life applications. In HGR, computers aim to identify and classify hand gestures. The limited diversity of the dataset used in HGR is due to the limited number of hand gesture demonstrators, acquisition environments, and hand pose variations despite previous efforts. Geometric image augmentations are commonly used to address these limitations. These augmentations include scaling, translation, rotation, flipping, and image shearing. However, research has yet to focus on identifying the best geometric transformations for augmenting the HGR dataset. This study employed three commonly utilized pre-trained models for image classification tasks, namely ResNet50, MobileNetV2, and InceptionV3. The system's performance was evaluated on five static HGR datasets: DLSI, HG14, ArabicASL, MU HandImages ASL, and Sebastian Marcell. The experimental results demonstrate that many geometric transformations are unnecessary for HGR image augmentation. Image shearing and horizontal flipping are the most influential transformations for augmenting the HGR dataset and achieving better classification performance. Moreover, ResNet50 outperforms MobileNetV2 and InceptionV3 for static HGR. [ABSTRACT FROM AUTHOR]

Recently, precision weed management has emerged as a promising solution for reducing the use of herbicides which is hazardous to crops and human health. Thus, accurate identification of weed in the early stage is the urge of current agricultural practices. Despite recent progress, developing an efficient weed identification system for real field scenarios is a serious challenge. To overcome this, a number of deep learning-based methods have been introduced in the literature. However, these methods require large volume of annotated data images which is rarely available. To address this gap, in this paper, a novel WeedGan method has been introduced for generating realistic synthetic images. The proposed WeedGan adopted concept of federated learning to reduce the computational load by introducing two discriminators. Additionally, a new loss function is defined for the efficient training of the generator. Extensive experiments have been performed to validate the performance of the WeedGan. First, the quality of images generated by the WeedGan was validated on cotton weed dataset in terms of FID score, and discriminator accuracy. The results are compared against four other state-of-the-art GAN models namely, DC-GAN, W-GAN, Info-Gan, and VIT-GAN. Further, classifiers performance of the generated dataset was evaluated using seven state-of-the-art transfer learning-based methods on the original, basic augmented, and WeedGan augmented datasets. The experimental results demonstrate that the proposed WeedGan has outperformed all the considered methods by achieving FID score of 282.76. Moreover, the classification performance of WeedGan augmented dataset was recorded highest with 97.82% on testing and 99.87% training accuracy with DenseNet121. [ABSTRACT FROM AUTHOR]

Data augmentation is a key technique to improve the generalizability of change detection models. Although the existing data augmentation methods achieve promising performance in image classification and object detection,they ignore the differences among the time-series images and the diversities of the changed objects. In order to preserve the change region and increase the information of the complex background, this paper proposed a novel data augmentation method for change detection based on change region mask, called MaskMix. Firstly, the change regions of the current image pair are pasted on an image pair to generate a new change image pair having new background and new changes. Secondly, MaskMix further augments the image pair by multi-path weighted fusion strategy. It selects a classical image processing method randomly from an image processing set for each path to conduct further augmentation. And then the processed image pairs from K paths are fused using a K-D weight generated by Dirichlet distribution. Finally, the pre-processed image pair and the post-processed image pair are fused by the weight generated by the Beta distribution through the skip connection. Experiments conducted on two publicly available datasets, e. g.,BCD（Build Change Detection） and LEVIR-CD（LEVIR Building Change Detection Dataset）, demonstrate that MaskMix significantly improves the generalizability of change detectors, e. g., ADCDNet, BIT, ChangeFormer,SNUNet, and DSAMNet. Moreover, compared with the existing image augmentation methods, such as MixUp, AugMix,MUM, and CropMix, MaskMix effectively increases the complexity and diversity of change images, enhancing the generalizability of existing change detection methods. [ABSTRACT FROM AUTHOR]