Efficient Deep Learning Model Research Articles

Lightweight and efficient deep learning models for fruit detection in orchards

The accurate recognition of apples in complex orchard environments is a fundamental aspect of the operation of automated picking equipment. This paper aims to investigate the influence of dense targets, occlusion, and the natural environment in practical application scenarios. To this end, it constructs a fruit dataset containing different scenarios and proposes a real-time lightweight detection network, ELD(Efficient Lightweight object Detector). The EGSS(Efficient Ghost-shuffle Slim module) module and MCAttention(Mix channel Attention) are proposed as innovative solutions to the problems of feature extraction and classification. The attention mechanism is employed to construct a novel feature extraction network, which effectively utilizes the low-latitude feature information, significantly enhances the fine-grained feature information and gradient flow of the model, and improves the model’s anti-interference ability. Eliminate redundant channels with SlimPAN to further compress the network and optimise functionality. The network as a whole employs the Shape-IOU loss function, which considers the influence of the bounding box itself, thereby enhancing the robustness of the model. Finally, the target detection accuracy is enhanced through the transfer of knowledge from the teacher’s network through knowledge distillation, while ensuring that the overall network is sufficiently lightweight. The experimental results demonstrate that the ELD network, designed for fruit detection, achieves an accuracy of 87.4%. It has a relatively low number of parameters (4.3×105\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$4.3 \ imes 10^5$$\\end{document}), a GLOPs of only 1.7, and a high FPS of 156. This network can achieve high accuracy while consuming fewer computational resources and performing better than other networks.

Deep Learning-Based Microscopic Damage Assessment of Fiber-Reinforced Polymer Composites.

Fiber-reinforced polymers (FRPs) are increasingly being used as substitutes for traditional metallic materials across various industries due to their exceptional strength-to-weight ratio. However, their orthotropic properties make them prone to multiple forms of damage, posing significant challenges in their design and application. During the design process, FRPs are subjected to various loading conditions to study their microscopic damage behavior, typically assessed through scanning electron microscopy (SEM). While SEM provides detailed insights into fracture surfaces, the manual analysis of these images is labor-intensive, time-consuming, and subject to variability based on the observer's expertise. To address these limitations, this research proposes a deep learning-based approach for the autonomous microscopic damage assessment of FRPs. Several computationally efficient pre-trained deep learning models, such as DenseNet121, NasNet Mobile, EfficientNet, and MobileNet, were evaluated for their performance in identifying different damage modes autonomously, thus reducing the need for manual interpretation. SEM images of FRPs with five distinct failure modes were used to validate the proposed method. These failure modes include three fiber-based failures such as fiber breakage, fiber pullout, and mixed-mode failure, and two matrix-based failures such as matrix brittle failure and matrix ductile failure. The entire dataset is divided into train, validation, and test sets. Deep learning models were established by training on train and validation sets for five failure modes, while the test set was used as the unseen data to validate the models. The models were assessed using various evaluation metrics on an unseen test dataset. Results indicate that the EfficientNet model achieved the highest accuracy of 97.75% in classifying the failure modes. The findings demonstrate the effectiveness of employing deep learning techniques for microscopic damage assessment, offering a more efficient, consistent, and scalable solution compared to traditional manual analysis.

A Deep Learning Model for Classifying Black Sigatoka Disease in Banana Leaves Based on Infection Stages

Objective: This research study aims to develop an efficient deep-learning model to detect and classify stages of Black Sigatoka disease in banana plants. Methods: In this study, deep learning techniques, specifically the basic Convolutional Neural Network (CNN) and VGG16 models, were used to address the challenge of identifying Black Sigatoka disease in banana leaves early on. The tests were conducted on a dataset containing labelled images of banana leaves, assessing their effectiveness based on criteria such as accuracy, precision, recall, and F1-score after adjusting hyperparameters for optimal outcomes. Findings: The results of the trials revealed that the basic CNN model attained a training accuracy of 96% and a validation accuracy of 89%, surpassing the performance of the VGG16 model. The VGG16 model, on the other hand, had a training accuracy of 92% and a validation accuracy of 89%. Across precision, recall, and F1 score measurements, the basic CNN model consistently outperformed the VGG16 model, with scores averaging 0.90 for all three metrics compared to VGG16's precision of 0.80, recall of 0.75, and F1 score of 0.75. The CNN model demonstrated its efficiency by stopping training at 26 epochs, whereas VGG16 completed training in 21 epochs. This demonstrates its effectiveness in detecting Black Sigatoka while utilising minimal resources. Novelty: A significant component of this study is its emphasis on identifying the stages of Black Sigatoka disease, which is commonly overlooked in research. By studying disease progression, this study provides insights for early intervention and disease management, aiding efforts to lessen the impact of Black Sigatoka on banana farming. Keywords: Black Sigatoka, Deep Learning, Disease Stages, Convolutional Neural Network, Classification, Identification

Computationally efficient deep learning models for diabetic retinopathy detection: a systematic literature review

Diabetic retinopathy, often resulting from conditions like diabetes and hypertension, is a leading cause of blindness globally. With diabetes affecting millions worldwide and anticipated to rise significantly, early detection becomes paramount. The survey scrutinizes existing literature, revealing a noticeable absence of consideration for computational complexity aspects in deep learning models. Notably, most researchers concentrate on employing deep learning models, and there is a lack of comprehensive surveys on the role of vision transformers in enhancing the efficiency of these models for DR detection. This study stands out by presenting a systematic review, exclusively considering 84 papers published in reputable academic journals to ensure a focus on mature research. The distinctive feature of this Systematic Literature Review (SLR) lies in its thorough investigation of computationally efficient approaches and models for DR detection. It sheds light on the incorporation of vision transformers into deep learning models, highlighting their significant contribution to improving accuracy. Moreover, the research outlines clear objectives related to the identified problem, giving rise to specific research questions. Following an assessment of relevant literature, data is extracted from digital archives. Additionally, in light of the results obtained from this SLR, a taxonomy for the detection of diabetic retinopathy has been presented. The study also highlights key research challenges and proposes potential avenues for further investigation in the field of detecting diabetic retinopathy.

Adaptive Hybrid Deep Learning Based Effective Channel Estimation in MIMO-Noma for Millimeter-Wave Systems with an Enhanced Optimization Algorithm

The channel estimation is crucial in the “millimeter Wave (mmWave) Massive Multiple-Input MultipleOutput (MIMO) and Non-Orthogonal Multiple Access (NOMA)” devices. Hybrid beamforming techniques are employed nowadays to minimize the complexity and equipment price. However, the absence of digital beam forming in mmWave affects the dynamic range and accuracy of the channel estimation. Previous research has concentrated mainly on predicting narrow-band mmWave channels using deep learning networks, as the wideband channels of mmWave create a considerable range and noise issue. Accurate channel estimation in the MIMO system is challenging because of the increased number of antennas and radio frequency (RF) chains. MIMO system communications using mmWave are frequently chosen because of their massive spectrum resources. Therefore, it is essential to tackle the obstacles obtained in the standard channel estimation framework by developing a MIMO-NOMA network with the help of deep learning methods Hence, this paper proposes an efficient hybrid deep learning model for channel estimation in MIMO-NOMA for mmWave systems. At first, the channel estimation is carried out using the Adaptive Hybrid Deep Learning (AHDL) model, where it combines both Autoencoder and Recurrent Neural Network (RNN). Here, the parameters are optimized using the Improved Red-tailed Hawk Algorithm (IRHA). Later the hardware cost and system complexity are reduced by performing the hybrid beam-forming process. Numerical results show that the proposed channel estimation and pilot estimation process outperforms the state-of-the-art approaches.

Combining 2.5D deep learning and conventional features in a joint model for the early detection of sICH expansion.

The study aims to investigate the potential of training efficient deep learning models by using 2.5D (2.5-Dimension) masks of sICH. Furthermore, it intends to evaluate and compare the predictive performance of a joint model incorporating four types of features with standalone 2.5D deep learning, radiomics, radiology, and clinical models for early expansion in sICH. A total of 254 sICH patients were enrolled retrospectively and divided into two groups according to whether the hematoma was enlarged or not. The 2.5D mask of sICH is constructed with the maximum axial, coronal and sagittal planes of the hematoma, which is used to train the deep learning model and extract deep learning features. Predictive models were built on clinic, radiology, radiomics and deep learning features separately and four type features jointly. The diagnostic performance of each model was measured using the receiver operating characteristic curve (AUC), Accuracy, Recall, F1 and decision curve analysis (DCA). The AUCs of the clinic model, radiology model, radiomics model, deep learning model, joint model, and nomogram model on the train set (training and Cross-validation) were 0.639, 0.682, 0.859, 0.807, 0.939, and 0.942, respectively, while the AUCs on the test set (external validation) were 0.680, 0.758, 0.802, 0.857, 0.929, and 0.926. Decision curve analysis showed that the joint model was superior to the other models and demonstrated good consistency between the predicted probability of early hematoma expansion and the actual occurrence probability. Our study demonstrates that the joint model is a more efficient and robust prediction model, as verified by multicenter data. This finding highlights the potential clinical utility of a multifactorial prediction model that integrates various data sources for prognostication in patients with intracerebral hemorrhage. The Critical Relevance Statement: Combining 2.5D deep learning features with clinic features, radiology markers, and radiomics signatures to establish a joint model enabling physicians to conduct better-individualized assessments the risk of early expansion of sICH.

Combining Transfer Learning and Ensemble Algorithms for Improved Citrus Leaf Disease Classification

Accurate categorization and timely control of leaf diseases are crucial for citrus growth. We proposed the Multi-Models Fusion Network (MMFN) for citrus leaf diseases detection based on model fusion and transfer learning. Compared to traditional methods, the algorithm (integrating transfer learning Alexnet, VGG, and Resnet) we proposed can address the issues of limited categories, slow processing speed, and low recognition accuracy. By constructing efficient deep learning models and training and optimizing them with a large dataset of citrus leaf images, we ensured the broad applicability and accuracy of citrus leaf disease detection, achieving high-precision classification. Herein, various deep learning algorithms, including original Alexnet, VGG, Resnet, and transfer learning versions Resnet34 (Pre_Resnet34) and Resnet50 (Pre_Resnet50) were also discussed and compared. The results demonstrated that the MMFN model achieved an average accuracy of 99.72% in distinguishing between diseased and healthy leaves. Additionally, the model attained an average accuracy of 98.68% in the classification of multiple diseases (citrus huanglongbing (HLB), greasy spot disease and citrus canker), insect pests (citrus leaf miner), and deficiency disease (zinc deficiency). These findings conclusively illustrate that deep learning model fusion networks combining transfer learning and integration algorithms can automatically extract image features, enhance the automation and accuracy of disease recognition, demonstrate the significant potential and application value in citrus leaf disease classification, and potentially drive the development of smart agriculture.

High-Precision Skin Disease Diagnosis through Deep Learning on Dermoscopic Images.

Dermatological conditions are primarily prevalent in humans and are primarily caused by environmental and climatic fluctuations, as well as various other reasons. Timely identification is the most effective remedy to avert minor ailments from escalating into severe conditions. Diagnosing skin illnesses is consistently challenging for health practitioners. Presently, they rely on conventional methods, such as examining the condition of the skin. State-of-the-art technologies can enhance the accuracy of skin disease diagnosis by utilizing data-driven approaches. This paper presents a Computer Assisted Diagnosis (CAD) framework that has been developed to detect skin illnesses at an early stage. We suggest a computationally efficient and lightweight deep learning model that utilizes a CNN architecture. We then do thorough experiments to compare the performance of shallow and deep learning models. The CNN model under consideration consists of seven convolutional layers and has obtained an accuracy of 87.64% when applied to three distinct disease categories. The studies were conducted using the International Skin Imaging Collaboration (ISIC) dataset, which exclusively consists of dermoscopic images. This study enhances the field of skin disease diagnostics by utilizing state-of-the-art technology, attaining exceptional levels of accuracy, and striving for efficiency improvements. The unique features and future considerations of this technology create opportunities for additional advancements in the automated diagnosis of skin diseases and tailored treatment.

Self-Attention-Based Convolutional Parallel Network: An Efficient Multi-Input Deep Learning Model for Endpoint Prediction of High-Carbon BOF Steelmaking

Self-Attention-Based Convolutional Parallel Network: An Efficient Multi-Input Deep Learning Model for Endpoint Prediction of High-Carbon BOF Steelmaking

Optimized quantum LSTM using modified electric Eel foraging optimization for real-world intelligence engineering systems

The integration of metaheuristics with machine learning methodologies presents significant advantages, particularly in optimization and computational intelligence. This amalgamation leverages the global search capabilities of metaheuristics alongside the pattern recognition and predictive prowess of machine learning, facilitating enhanced convergence rates and solution quality in complex problem spaces. The Quantum Long Short-Term Memory (QLSTM) emerges as a highly efficient deep learning model tailored to tackle such intricate engineering problems. The QLSTM's architecture, comprising data encoding, variational, and quantum measurement layers, facilitates the effective encoding and processing of civil engineering data, leading to heightened prediction accuracy. However, the task of determining optimal values for QLSTM parameters presents challenges due to its NP-problem nature and time-consuming characteristics. To address this, we propose an alternative technique to optimize the QLSTM based on a modified Electric Eel Foraging Optimization (MEEFO). The MEEFO is a modified version of the original EEFO that applies triangular mutation operators to boost the search capability of the traditional EEFO. Thus, the MEEFO optimizes the QLSTM and boosts its prediction performance. To validate the efficacy of our proposed method, we conduct comprehensive experiments utilizing five real-world engineering datasets related to construction and structure engineering. The evaluation outcomes unequivocally demonstrate that the MMEFO significantly enhances the performance of the QLSTM.

Wasserstein generative adversarial network with gradient penalty and convolutional neural network based motor imagery EEG classification

Objective.Due to the difficulty in acquiring motor imagery electroencephalography (MI-EEG) data and ensuring its quality, insufficient training data often leads to overfitting and inadequate generalization capabilities of deep learning-based classification networks. Therefore, we propose a novel data augmentation method and deep learning classification model to enhance the decoding performance of MI-EEG further.Approach.The raw EEG signals were transformed into the time-frequency maps as the input to the model by continuous wavelet transform. An improved Wasserstein generative adversarial network with gradient penalty data augmentation method was proposed, effectively expanding the dataset used for model training. Additionally, a concise and efficient deep learning model was designed to improve decoding performance further.Main results.It has been demonstrated through validation by multiple data evaluation methods that the proposed generative network can generate more realistic data. Experimental results on the BCI Competition IV 2a and 2b datasets and the actual collected dataset show that classification accuracies are 83.4%, 89.1% and 73.3%, and Kappa values are 0.779, 0.782 and 0.644, respectively. The results indicate that the proposed model outperforms state-of-the-art methods.Significance.Experimental results demonstrate that this method effectively enhances MI-EEG data, mitigates overfitting in classification networks, improves MI classification accuracy, and holds positive implications for MI tasks.

Design and implementation of efficient distributed deep learning model inference architecture on serverless computation

Although the distributed algorithm and architecture have been widely used in the development of traditional large-scale machine model reasoning tasks, but it is unavoidable about the source waste caused by synchronization between learners and participants. However, serverless computing is charged based on usage, which makes it a popular alternative for model inference. This article mainly proposes the architecture and algorithm that can be used for the inference task of the distributed deep learning model and adopts coarse-grained parallelization strategy to solve the problem of high communication cost between Serverless platform functions. The paper tested the performance of the model on the deep learning model and found that the model reasoning based on the Serverless platform can better cope with the requirements and changes, and ensure the stability of the system. It is worth noting that both the data owner and the model owner can benefit from the process of carrying out the machine learning model inference task, but existing solutions cannot satisfied with them privacy requirements, which is a problem needs to think about in the field of machine learning inference tasksbased on the Serverless platform in the future.

Deep learning in high-frequency trading: Conceptual challenges and solutions for real-time fraud detection

High-frequency trading (HFT) has transformed financial markets by enabling rapid execution of trades, exploiting market inefficiencies, and optimizing trading strategies. However, this speed and complexity also present significant challenges for real-time fraud detection. Deep learning, a subset of machine learning, offers promising solutions to these challenges through its ability to analyze large volumes of data and uncover intricate patterns. This review explores the conceptual challenges and solutions associated with deploying deep learning for fraud detection in HFT environments. One of the primary challenges in implementing deep learning for HFT fraud detection is the sheer volume and velocity of data. HFT systems generate vast amounts of transactional data in milliseconds, necessitating highly efficient and scalable deep learning models. Convolutional neural networks (CNNs) and recurrent neural networks (RNNs) are particularly suited for this task due to their ability to process and analyze sequential data efficiently. However, these models require substantial computational resources and sophisticated infrastructure to operate in real time. Another significant challenge is the need for high accuracy and low latency in fraud detection. False positives can lead to unnecessary interventions, while false negatives can result in undetected fraudulent activities. Deep learning models must be fine-tuned to balance these risks, employing techniques such as hyperparameter optimization and ensemble learning to enhance their predictive capabilities. Additionally, integrating real-time anomaly detection methods can help identify suspicious activities promptly, reducing the window of opportunity for fraudsters. Data quality and integrity also pose substantial challenges. HFT environments are susceptible to noise and outliers, which can distort model predictions. Ensuring high-quality data through rigorous preprocessing and anomaly filtering is crucial for the accuracy of deep learning models. Techniques such as data augmentation and normalization can further improve model robustness. To address these challenges, a hybrid approach combining deep learning with traditional statistical methods and rule-based systems can be effective. This approach leverages the strengths of each method, providing a comprehensive fraud detection framework that is both accurate and responsive. Additionally, ongoing model retraining and adaptation to evolving fraud patterns are essential to maintain the effectiveness of the system. In conclusion, while deep learning presents significant opportunities for enhancing real-time fraud detection in high-frequency trading, it also requires addressing challenges related to data volume, computational demands, accuracy, and data quality. By employing a hybrid approach and continually refining models, financial institutions can effectively mitigate fraud risks and safeguard their trading operations.

A new paradigm in cigarette smoke detection: Rapid identification technique based on ATR-FTIR spectroscopy and GhostNet-[formula omitted]

In the tobacco field, the analysis and classification of cigarette smoke are crucial for ensuring product quality, protecting consumer health, and driving industry innovation. These analyses provide a comprehensive understanding of tobacco products and serve as a scientific basis for production standards and regulatory policies. Traditional infrared spectroscopy techniques for tobacco quality analysis often suffer from limitations such as time consumption, high costs, and complexity. To achieve efficient detection of cigarette smoke and quality control of cigarettes, this study developed a cigarette mainstream smoke detection system based on Attenuated Total Reflectance Fourier transform infrared (ATR-FTIR) spectroscopy technology, aimed at rapidly acquiring the mid-infrared (MIR) spectral characteristics of cigarette smoke and combining deep learning for rapid cigarette type identification. Five different brands of cigarettes were selected for this study, and a total of 400 MIR spectra of cigarette smoke aerosols were collected. Using three spectral preprocessing methods and two-dimensional correlation spectroscopy (2DCOS), we extracted the fingerprint region (1850–650 cm-1) to generate 4800 2DCOS images. We then optimized the GhostNet network architecture and constructed a more compact and efficient deep learning model, GhostNet-α. A mixed-precision training strategy further enhanced model training speed and computational efficiency. Results show that Multiple Scatter Correction (MSC)-preprocessed synchronous 2DCOS images achieved 100 % classification accuracy. This study validates GhostNet-α’s efficiency and accuracy in cigarette smoke recognition, offering a new technical approach for quality control and rapid detection in the tobacco industry, significantly enhancing its competitiveness and consumer health protection.

ST-TrajGAN: A synthetic trajectory generation algorithm for privacy preservation

The rapid growth of large-scale trajectory data poses privacy risks for location-based services (LBS), primarily through centralized storage and processing of data, as well as insecure data transmission channels (such as the Internet and wireless networks), which can lead to unauthorized access or manipulation of users' location information by attackers. To enhance trajectory privacy protection while improving the trajectory utility, this paper proposes an efficient and secure deep learning model Semantic and Transformer-based Trajectory Generative Adversarial Networks (ST-TrajGAN) for trajectory data generation and publication. First, this article introduces a semantic trajectory encoding model for preprocessing trajectory points. Through this model, trajectory points can be transformed into vector representations with semantic information. Next, by learning the spatio-temporal and semantic features of real trajectory data, a deep learning model is used to generate synthetic trajectories with more uncertainty and practicality. Furthermore, a novel TrajLoss loss metric function was crafted to gauge the trajectory similarity loss within the trained deep learning model. Ultimately, the efficacy of the generated synthetic trajectories and the model's utility are assessed through Trajectory-User Linking (TUL) and Trajectory Sharing Percentage (TSP) values on three authentic Location-Based Services (LBS) datasets. Numerous experiments have shown that our method outperforms other methods in terms of privacy protection effectiveness and utility.

Analysis of RNA translation with a deep learning architecture provides new insight into translation control.

Accurate annotation of coding regions in RNAs is essential for understanding gene translation. We developed a deep neural network to directly predict and analyze translation initiation and termination sites from RNA sequences. Trained with human transcripts, our model learned hidden rules of translation control and achieved a near perfect prediction of canonical translation sites across entire human transcriptome. Surprisingly, this model revealed a new role of codon usage in regulating translation termination, which was experimentally validated. We also identified thousands of new open reading frames in mRNAs or lncRNAs, some of which were confirmed experimentally. The model trained with human mRNAs achieved high prediction accuracy of canonical translation sites in all eukaryotes and good prediction in polycistronic transcripts from prokaryotes or RNA viruses, suggesting a high degree of conservation in translation control. Collectively, we present a general and efficient deep learning model for RNA translation, generating new insights into the complexity of translation regulation.

Non-Intrusive Load Monitoring Based on an Efficient Deep Learning Model With Local Feature Extraction

Non-Intrusive Load Monitoring Based on an Efficient Deep Learning Model With Local Feature Extraction

Efficient deep-learning model for predicting mechanical strength of carbon nanotube-reinforced cementitious composites: Database acquisition, framework development and implementation

In response to the growing demand for accurate models predicting the mechanical strength of carbon nanotube (CNT) reinforced cementitious composites, this study introduces a precise deep-learning model. The model, built upon a comprehensive dataset of 219 entries, incorporates 14 input variables considering various composites, cement types, and CNT characteristics. Notably, analysis of the data demonstrates a nuanced understanding of material properties, covering cementitious composites with compressive strength (CS) and flexural strength (FS) ranging from 15 to 140 MPa and 3–17 MPa, respectively. Employing a regularized deep learning neural network architecture, the model exhibits promising performance metrics, with the training dataset showcasing the lowest mean absolute error (MAE) of 2.036. The training and validation datasets demonstrate a high correlation coefficient (0.991 and 0.917, respectively). However, higher MAE values are observed for the testing and validation datasets, indicating potential limitations in generalization. Furthermore, the model's robustness is highlighted by its ability to surpass earlier studies in accuracy for validation datasets. The study elucidates a nonlinear trend in FS and the influence of CNT size on CS, with implications for material design and construction practices.

Deep-Learning-Based Automatic Sinkhole Recognition: Application to the Eastern Dead Sea

Sinkholes can cause significant damage to infrastructures, agriculture, and endanger lives in active karst regions like the Dead Sea’s eastern shore at Ghor Al-Haditha. The common sinkhole mapping methods often require costly high-resolution data and manual, time-consuming expert analysis. This study introduces an efficient deep learning model designed to improve sinkhole mapping using accessible satellite imagery, which could enhance management practices related to sinkholes and other geohazards in evaporite karst regions. The developed AI system is centered around the U-Net architecture. The model was initially trained on a high-resolution drone dataset (0.1 m GSD, phase I), covering 250 sinkhole instances. Subsequently, it was additionally fine-tuned on a larger dataset from a Pleiades Neo satellite image (0.3 m GSD, phase II) with 1038 instances. The training process involved an automated image-processing workflow and strategic layer freezing and unfreezing to adapt the model to different input scales and resolutions. We show the usefulness of initial layer features learned on drone data, for the coarser, more readily-available satellite inputs. The validation revealed high detection accuracy for sinkholes, with phase I achieving a recall of 96.79% and an F1 score of 97.08%, and phase II reaching a recall of 92.06% and an F1 score of 91.23%. These results confirm the model’s accuracy and its capability to maintain high performance across varying resolutions. Our findings highlight the potential of using RGB visual bands for sinkhole detection across different karst environments. This approach provides a scalable, cost-effective solution for continuous mapping, monitoring, and risk mitigation related to sinkhole hazards. The developed system is not limited only to sinkholes however, and can be naturally extended to other geohazards as well. Moreover, since it currently uses U-Net as a backbone, the system can be extended to incorporate super-resolution techniques, leveraging U-Net based latent diffusion models to address the smaller-scale, ambiguous geo-structures that are often found in geoscientific data.

SKINC-NET: an efficient Lightweight Deep Learning Model for Multiclass skin lesion classification in dermoscopic images

SKINC-NET: an efficient Lightweight Deep Learning Model for Multiclass skin lesion classification in dermoscopic images