Deep Learning Mechanisms Research Articles

An Efficient Deep Learning Mechanism for Predicting Fake News/Reviews in Twitter Data

Recently, social media platforms have been widely utilized as information sources due to their effortless accessibility and reduced costs. However, online platforms like Instagram, Twitter and Facebook get influenced by their users via fake news/reviews. The main intention of spreading fake news is to mislead other network users, which highly affects businesses, political parties, etc. Thus, an effective methodology is needed to predict fake news from social media automatically. The major objective of this proposed study is to identify and classify the given Twitter input data as real or fake through deep learning mechanisms. The proposed study involves four stages: pre-processing, embedded word analysis, feature extraction, and fake news/reviews prediction. Initially, pre-processing is performed to enhance the quality of data with the help of tokenization, stemming and stop word removal. Embedded word analysis is done using Advanced Word2Vec and GloVe modeling to enhance the performance of a proposed prediction model. Then, the hybrid deep learning model named Dense Convolutional assisted Gannet Optimal Bi-directional Network (DC_GO_BiNet) is introduced for feature extraction and prediction. A Dense Convolutional Neural Network (DCNN) is hybridized with a bi-directional long-short-term memory (Bi-LSTM) model to extract the essential features and predict fake news from the given input text. Also, the proposed model’s parameters are fine-tuned by adopting a gannet optimization (GO) algorithm. The proposed study used three different datasets and obtained higher classification accuracy as 99.5% in Fake News Detection on Twitter EDA, 99.59% in FakeNewsNet and 99.51% in ISOT. The analysis proves that the proposed model attains higher prediction results for each dataset than others.

Stress-induced overeating behaviors explained from a (transitory) relief-learning perspective

People use various behaviors to cope with stressful events. These behaviors are mostly adaptive, as they allow a successful release of stress without impacting other aspects of life: How nice is it to have a break with a few spoons of that favorite ice cream during a hectic working day? However, when excessive consumption of high-sugar/salt ultra-processed food becomes the gateway to find relief from stress, eating loses this adaptive function and may escalate to binge eating, lead to obesity, and other medical conditions linked to overweight.Several etiological models attempt to explain stress-induced eating and excessive overeating behaviors characterizing these clinical conditions. The popular Emotional Eating Theory proposes that stress-related (over-)eating, a major predictor of obesity and diagnosed binge eating disorders, develops based on negative reinforcement learning since food consumption regulates the negative affective state associated with stressful circumstances. Differently, the prominent Incentive Sensitization Theory explains overeating, binge eating disorders (including bulimia), and obesity in terms of excessive amplification of reward ‘wanting’, which is thought to emerge from overexposure to obesogenic (food)cues. The several studies oriented by these theories have paved the way to better understand stress-related (over-)eating and its clinical excesses. However, a deep mechanistic understanding of how and why stress-induced (over-)eating can escalate till clinical forms of overeating remain elusive. A well-funded connection of the mechanisms proposed by the Emotional Eating Theory and the Incentive Sensitization Theory might address this etiological open question. To avoid erroneous arguments, it is however essential to first address the internal theoretical and methodological shortcomings of each theory and connected studies. These shortcomings stem from conceptual fallacies and poorly implemented designs, which might partially explain the ‘high variability and low replicability’ problem of empirical findings. Next, the formulation of a new integrative model could provide fresh insight into the deep learning and biological mechanisms of this escalation. A successful formalization of this model could then create the much-needed impact in clinical and preventive research since excessive overeating is a behavior hard to change once established.In this opinion paper, I propose to apply recent insights we gathered on the role of relief from the field of safety learning to stress (over-)eating. I will present a new relief-based model that, as a starting point, has the potential to connect the Emotional Eating Theory with the Incentive Sensitization Theory, setting the base for more integrative science.

Advancing in RGB-D Salient Object Detection: A Survey

The human visual system can rapidly focus on prominent objects in complex scenes, significantly enhancing information processing efficiency. Salient object detection (SOD) mimics this biological ability, aiming to identify and segment the most prominent regions or objects in images or videos. This reduces the amount of data needed to process while enhancing the accuracy and efficiency of information extraction. In recent years, SOD has made significant progress in many areas such as deep learning, multi-modal fusion, and attention mechanisms. Additionally, it has expanded in real-time detection, weakly supervised learning, and cross-domain applications. Depth images can provide three-dimensional structural information of a scene, aiding in a more accurate understanding of object shapes and distances. In SOD tasks, depth images enhance detection accuracy and robustness by providing additional geometric information. This additional information is particularly crucial in complex scenes and occlusion situations. This survey reviews the substantial advancements in the field of RGB-Depth SOD, with a focus on the critical roles played by attention mechanisms and cross-modal fusion methods. It summarizes the existing literature, provides a brief overview of mainstream datasets and evaluation metrics, and quantitatively compares the discussed models.

PI-STGnet: Physics-integrated spatiotemporal graph neural network with fundamental diagram learner for highway traffic flow prediction

At present, traffic state prediction primarily relies on purely data-driven methods, ignoring the incorporation of physical constraints within the field of traffic flow. Taking this as a starting point, this paper endeavors to embed the physical mechanism of the traffic flow fundamental graph into the deep learning model, and proposes a physics-integrated spatiotemporal graph neural network with fundamental diagram learner (PI-STGnet) for highway traffic flow prediction. The PI-STGnet mainly comprises a semantic enhancement module (SE), a spatiotemporal correlation extraction module (ST-Block), and a fundamental diagram learner module (FD-learner). These modules are strategically designed to sequentially enhance the contextual semantic relationships of the traffic flow and speed inputs, extract intricate spatiotemporal correlations of traffic states, and adaptively learn the dynamic evolution of the physical relationships between traffic flow and speed. The real-world dataset employed to assess the performance of the PI-STGnet is sourced from the monitoring data of highway gantry sensors located in Ningde City, Fujian Province, China. The experimental results demonstrate that the proposed PI-STGnet has the advantages to extract spatiotemporal features and achieve prediction, surpassing the accuracy of cutting-edge baseline models. In summary, as an exploratory work, this paper provides a novel approach that driven by the fusion of deep learning and physical mechanisms for traffic flow prediction to a certain extent.

Precise Adverse Weather Characterization by Deep-Learning-Based Noise Processing in Automotive LiDAR Sensors

With current advances in automated driving, optical sensors like cameras and LiDARs are playing an increasingly important role in modern driver assistance systems. However, these sensors face challenges from adverse weather effects like fog and precipitation, which significantly degrade the sensor performance due to scattering effects in its optical path. Consequently, major efforts are being made to understand, model, and mitigate these effects. In this work, the reverse research question is investigated, demonstrating that these measurement effects can be exploited to predict occurring weather conditions by using state-of-the-art deep learning mechanisms. In order to do so, a variety of models have been developed and trained on a recorded multiseason dataset and benchmarked with respect to performance, model size, and required computational resources, showing that especially modern vision transformers achieve remarkable results in distinguishing up to 15 precipitation classes with an accuracy of 84.41% and predicting the corresponding precipitation rate with a mean absolute error of less than 0.47 mm/h, solely based on measurement noise. Therefore, this research may contribute to a cost-effective solution for characterizing precipitation with a commercial Flash LiDAR sensor, which can be implemented as a lightweight vehicle software feature to issue advanced driver warnings, adapt driving dynamics, or serve as a data quality measure for adaptive data preprocessing and fusion.

Modelling and Analysis of Smart Tourism Based on Deep Learning and Attention Mechanism

In the current traditional tourism recommendation systems, significant amounts of manpower and resources are required to manually identify the characteristics of resources, resulting in extremely poor economic benefits. To address this issue, this study proposes a smart tourism model based on deep learning and attention mechanisms. It uses a deep learning model to extract semantic information and improves it with the attention mechanism. This is to enable the model to take into account the complete meaning of the text and the association between individual words, thereby achieving a more comprehensive extraction of tourism resource features. The experiment showcases that the [Formula: see text]1-value of the algorithm proposed by us reached 0.961, the Recall value reached 0.958, the accuracy reached 0.980 and the area under the receiver operating characteristic curve reached 0.956. All parameters are superior to the comparison algorithm, and in practical application testing, its fitting degree reached 0.981. The above results indicate that the smart tourism proposed by us based on deep learning and attention mechanism has excellent performance in the field of tourism resource recommendation, which can effectively extract hidden features from the resources. This can also accurately push the tourism resources that users are interested in, which can effectively promote the integration and development of the tourism industry and the Internet, and has strong positive significance for economic development.

An Overview of Behavioral Recognition

Human behavior recognition has become a popular research topic in the field of computer vision. With the introduction of deep learning and attention mechanisms, this field has been further promoted. However, issues such as dataset acquisition and preprocessing operations on multimodal datasets, modeling of long time information in videos, and fusion of temporal and spatial information still exist. In this paper, we first outline some video action recognition datasets and related preprocessing techniques, including frame extraction, optical flow extraction, and skeletal feature acquisition. Then, the relevant models are classified and parsed according to their characteristics and the types of input data modalities. In addition, we evaluate the performance of the models on several benchmark datasets to gain a deeper understanding of the model development process. Finally, we summarized the current challenges faced in the field of video behavior recognition, including model timeliness, data set subjectivity and effective fusion of multi-modal features, and proposed possible future improvement directions in order to provide more ideas and methods for subsequent research.

A hybrid CNN-GRU-based intrusion detection system for secure communication in vehicular adhoc network

ABSTRACT Vehicular Ad hoc Network (VANET) is a component of the Intelligent Transportation System (ITS) which furnishes communication among vehicles. It delivers comfort and safety information to passengers and vehicle drivers. Security plays a vital role during the transmission of data as the various distinct security attacks directly influence the safety of the passengers on the road. Several security attacks will disrupt normal functions like the transmission of data. Some security breaches inject false information which affects the safety of drivers. However, there are several demurrers in detecting abnormal activities efficiently as the traditional system faces imbalanced data problems. This paper presents an Intrusion Detection System (IDS) for detecting anomalies and protecting communication systems from several distinct attacks. The proposed IDS utilizes two deep learning (DL) mechanisms such as Convolutional Neural Network (CNN) and Gated Recurrent Unit (GRU) for detecting anomalies in the vehicular network. The hybrid CNN-GRU mechanism helps in solving the overfitting and low-velocity problems. It also detects several attacks and protects the ad hoc network from those attacks to safeguard the vehicle users. The results show that the propounded model outperforms the other traditional detection mechanisms by 10.79% in terms of performance compared to other IDS.

Security protocols in healthcare: A comprehensive study of AI-Enabled IoMT

The Internet of Medical Things IoMT has driven a paradigm shift in the way things have been done traditionally within medical Healthcare practices and this is evidence that Smart Personal systems (Health Care) are shaping up with mighty forces. By merging these technologies, IoT has transformed data communication and provided the means to directly interconnect medical devices and sensor systems. The use of advanced technologies like Artificial Intelligence( AI), Machine Learning, Deep Learning, and Federated Learning has highly equipped and transformed the IoMT’s performance. These integrations also raise concerns on data privacy and security, and addressing these challenges become more important as without proper security and privacy measures people can face serious disruption both online and offline leading to upcoming catastrophes already waiting at its end line. This chapter looks into the intricacies of a highly secure communications system powered by AI, specifically in the age where disasters like COVID-19 [10], emphasizes the need for increased security on data transmission as well as storage. IoMT has wide applications for integrating remote patient monitoring, enhancing user satisfaction and interdimensional hospitalization rates, we delve into how IoMT works in conjunction with machine learning, particularly federated Learning intends to address any possible concerns. The chapter begins with IoMT systems, including IoMT ecosystems, implantable medical devices IMDs, and Internet-worn gadgetry. Special attention is given to the AI-powered Intrusion Detection Systems (IDSs) in IoMT environments, while network security remains the research focus helping practitioners to implement safety measures enterprise-wide; analysis, comparison, and evaluation all encompass machine learning, deep learning mechanisms as precautions against potential attacks. We will also consider another study on the services of an AI-provisioned lightweight protocol communication system that ensures data authenticity and responds to diverse security implications in IMD. The chapter makes a U-turn by alluding to internet wearable devices, pointing out federated learning and blockchain reinforcements that indicate the need for an integrated framework. The amalgamation of future technologies supports the ethical and secure integration of IoMT. Collaboration, as a guiding principle, compels healthcare practitioners and policymakers to partner efficiently to guide the future where the security of health is ever easier. In the end, a summary of modern ethical and regulatory paradigms in IoMT is provided, emphasizing that there could be an urgent need for new laws.

Deep Learning for Predicting Progression of Patellofemoral Osteoarthritis Based on Lateral Knee Radiographs, Demographic Data, and Symptomatic Assessments.

In this study, we propose a novel framework that utilizes deep learning and attention mechanisms to predict the radiographic progression of patellofemoral osteoarthritis (PFOA) over a period of 7 years. This study included subjects (1,832 subjects, 3,276 knees) from the baseline of the Multicenter Osteoarthritis Study (MOST). Patellofemoral joint regions of interest were identified using an automated landmark detection tool (BoneFinder) on lateral knee X-rays. An end-to-end deep learning method was developed for predicting PFOA progression based on imaging data in a five-fold cross-validation setting. To evaluate the performance of the models, a set of baselines based on known risk factors were developed and analyzed using gradient boosting machine (GBM). Risk factors included age, sex, body mass index, and Western Ontario and McMaster Universities Arthritis Index score, and the radiographic osteoarthritis stage of the tibiofemoral joint (Kellgren and Lawrence [KL] score). Finally, to increase predictive power, we trained an ensemble model using both imaging and clinical data. Among the individual models, the performance of our deep convolutional neural network attention model achieved the best performance with an area under the receiver operating characteristic curve (AUC) of 0.856 and average precision (AP) of 0.431, slightly outperforming the deep learning approach without attention (AUC = 0.832, AP = 0.4) and the best performing reference GBM model (AUC = 0.767, AP = 0.334). The inclusion of imaging data and clinical variables in an ensemble model allowed statistically more powerful prediction of PFOA progression (AUC = 0.865, AP = 0.447), although the clinical significance of this minor performance gain remains unknown. The spatial attention module improved the predictive performance of the backbone model, and the visual interpretation of attention maps focused on the joint space and the regions where osteophytes typically occur. This study demonstrated the potential of machine learning models to predict the progression of PFOA using imaging and clinical variables. These models could be used to identify patients who are at high risk of progression and prioritize them for new treatments. However, even though the accuracy of the models were excellent in this study using the MOST dataset, they should be still validated using external patient cohorts in the future.

A Comparison of Customer Churn Vector Embedding Models with Deep Learning

In the telecommunication industry, Deep learning has been utilized for churn prediction. Some companies have used sophisticated deep learning techniques to predict churn, which yielded good results. However, future studies are still required to evaluate several deep learning mechanisms as only SoftMax Loss has been used so far. By comparing customer churn vector embedding models with several methods, including SoftMax Loss, Large Margin Cosine Loss, Semi-Supervised Learning, and a combination of Large Margin Cosine Loss and Semi-Supervised Learning, we continue our previous research to apply deep learning in predicting customer churn in the telecommunications industry in this paper. The use of Large Margin Cosine Loss has been proven in face recognition which can increase the discrimination between vectors embedding in different classes. Understanding how mixing unlabeled and labeled input might alter developing algorithms and learning behavior that benefit from this combination are the goals of semi-supervised learning. This approach successfully encouraged feature discrimination in customer behavior as well as improved the overall accuracy of the model. Large Margin Cosine Loss in this study achieved 83.74% of the F1 Score compared to other methods. It was further demonstrated that the produced vectors for churn prediction are discriminative by examining the cluster's similarity and the t-SNE plot.

An efficient deep learning mechanisms for IoT/Non-IoT devices classification and attack detection in SDN-enabled smart environment

In recent years, the development of Internet of Things (IoT) applications has increased, resulting in higher demands for sufficient bandwidth, data rates, latency, and quality of service (QoS). In advanced communications, managing network resources for allocating IoT services and identifying the exact IoT devices connected to a network is a major concern. The existing studies have introduced various methods for classifying IoT devices in a network. However, the previous studies faced challenges like limited attributes, low efficiency, inappropriate features, and computational complexities. Also, the existing studies failed to concentrate on IoT/Non-IoT classification along with attack detection. Detecting attacks on IoT devices is critical for making network services more effective. Thus, the proposed study introduces an efficient IoT device classification and attack detection mechanism using software defined networking (SDN)-enabled fiber-wireless access networks internet of things (FiWi IoT) architecture. Initially, an effective resource allocation process is performed to mitigate the delay constraint issues by introducing a hybrid parallel neural network-based dynamic bandwidth allocation (DBA) method. Then, the input traffic information is gathered from the resource-efficient SDN-enabled FiWi IoT network, and the input data is pre-processed to eliminate unwanted noises using min-max normalization and standardization. Next, the essential attributes are extracted to attain enhanced classification performance. To reduce the feature dimensionality problem and thereby solve complexity issues, the most optimal features are selected by a new chaotic seagull optimization (CSO) approach. After that, IoT/non-IoT classification is performed using a transformer-driven deep intelligent model. Finally, the attacks are detected and classified by introducing a novel slice attention-based deep capsule autoencoder (SA_DCAE) model. For experimentation, the Python 3.7.0 tool is used in this work, and the performance of proposed classifiers is measured by evaluating varied matrices. Also, the comparison analysis proves the superiority of the proposed techniques to other existing methods.

Secure cloud storage for IoT based distributed healthcare environment using blockchain orchestrated and deep learning model

Internet of Things (IoT), a distributed healthcare system has integrated different medical resources with sensors and actuators. In this research paper proposes a secure healthcare monitoring system for IoT based distributed healthcare systems in the cloud using blockchain and deep learning (DL) mechanisms. The proposed system involved three phases: secure data transmission, data storage, and disease classification system. Initially, the patients are authenticated via blockchain mechanism and their data is encrypted via Effective Key-based Rivest Shamir Adelman (EKRSA), in which the keys are generated using Circle chaotic map and Linear inertia weight-based Honey Badger Optimization (CLHBO) algorithm. Next, in the data storage phase, these encrypted IoT data are securely stored in the cloud using blockchain technology in a distributed manner. Finally, in the disease classification, the data are gathered from the publicly available dataset, and these collected datasets are preprocessed to handle missing values and data normalization. After that, the proposed system applies a radial basis kernel-based linear discriminant analysis (RBKLDA) model to reduce the dimensionality of the dataset. At last, the disease classification is done by optimal parameter-centered bidirectional long short-term memory (OPCBLSTM). The proposed EKRSA system archives maximum throughput of 99.05% and reliability of 99.66, which is superior to the existing approaches. The OPCBLSTM is investigated for its disease classification process, the proposed one achieves 99.64% accuracy with less processing time of 6 ms, which is superior to the existing classifiers. The experimental analysis proves that the system attained better security and classification metrics results than the existing methods.

IMPROVING DETECTION FOR INTRUSION USING DEEP LSTM WITH HYBRID FEATURE SELECTION METHOD

Due to the importance of the intrusion detection system, which is considered supportive of enhancing network security. Therefore, we seek to increase the efficiency of intrusion detection systems through the use of deep learning mechanisms. However, intrusion detection algorithms still suffer from problems in the process of classification and determining the presence and type of attack, which causes a decrease in the detection rate, an increase in the number of false alarms, and reduces system performance. This is due to a large number of redundant features that are not relevant to the dataset. To find the solve for this problem, here propose a hybrid algorithm based on the use of the feature selection technique, which helps in reaching the goal optimally by choosing the best and most important features. And it works by integrating three ways to minimize the features by deleting the static features, that do not have much value from the information gain and is done before the training stage by the deep learning model of LSTM as preprocessing for the CSE-CIC-IDS data set, which helps in improving the performance of the system By minimizing the time for processing and increasing the detection rate and accuracy ratio. The results of the experiment showed a high accuracy of 99.84%.

TemBERTure: advancing protein thermostability prediction with deep learning and attention mechanisms.

Understanding protein thermostability is essential for numerous biotechnological applications, but traditional experimental methods are time-consuming, expensive, and error-prone. Recently, deep learning (DL) techniques from natural language processing (NLP) was extended to the field of biology, since the primary sequence of proteins can be viewed as a string of amino acids that follow a physicochemical grammar. In this study, we developed TemBERTure, a DL framework that predicts thermostability class and melting temperature from protein sequences. Our findings emphasize the importance of data diversity for training robust models, especially by including sequences from a wider range of organisms. Additionally, we suggest using attention scores from Deep Learning models to gain deeper insights into protein thermostability. Analyzing these scores in conjunction with the 3D protein structure can enhance understanding of the complex interactions among amino acid properties, their positioning, and the surrounding microenvironment. By addressing the limitations of current prediction methods and introducing new exploration avenues, this research paves the way for more accurate and informative protein thermostability predictions, ultimately accelerating advancements in protein engineering. TemBERTure model and the data are available at: https://github.com/ibmm-unibe-ch/TemBERTure.

Community Microgrid Energy Co-Scheduling Based on Deep Reinforcement Learning and Contribution Mechanisms

Community Microgrid Energy Co-Scheduling Based on Deep Reinforcement Learning and Contribution Mechanisms

Image‐based detection and classification of plant diseases using deep learning: State‐of‐the‐art review

AbstractPlant diseases are assumed to be one of the primary causes regulating food manufacturing and reducing deficits in crop yield, and it is crucial that plant diseases have rapid spotting and identification. Since the demand for food is rising quickly due to population growth, the main issue for any country is plant disease mechanization in agricultural science. Recent developments in deep learning techniques have found use in plant disease recognition, offering a robust tool with exceedingly precise results. Our main goal is to employ deep learning mechanisms to better understand diseases and how to identify them quickly. In this regard, we analyzed 94 publications chosen from the last 7 years (2016–2023) that show how CNN's philosophy has evolved over this period with various approaches to treating plant diseases. Furthermore, a full description of numerous crops, diseases connected to them, various datasets relating to plant diseases, existing CNN models, and customized CNN architectures is provided. The results of this state‐of‐the‐art review can be implemented to comprehend the cutting‐edge trends in the application of deep learning (CNNs) to detect plant diseases as well as pinpoint any research gaps that require the scientific community's attention.

Does the face expression recognition model under the online entrepreneurial education model important for deep learning technologies and entrepreneurial emotions?

AI's rapid development has already affected the economy and society, and it might revolutionize how goods and services are manufactured and categorized. This affects the country's production, employment rate, and competitiveness. This research aims to determine how emotions (good and negative) impact deep learning technologies and company startup intentions. The research also examined how emotion recognition influences online entrepreneurship education and how deep learning and entrepreneurship are connected. In conclusion, the research is one of the pioneering efforts that provides a comprehensive grasp of deep learning and facial recognition in entrepreneurship. 78 of the 650 replies were insufficient, according to the statistics. Some individuals constantly picked 3, while others submitted blank questionnaires. The answer is inadequate. 572 surveys were utilized for the study. Positive emotions affect the urge to establish a company, according to the research. People who are generally happy may be able to develop their entrepreneurial expertise and be more driven to behave entrepreneurially. Positive and negative emotions have a significant impact on the deep learning technology and entrepreneurial intention connection. The findings imply that individuals who often experience pleasant emotions may be better able to improve their entrepreneurial cognition and, as a consequence, their drive to engage in entrepreneurial activity. The facial recognition model in entrepreneurship significantly moderates the relationship between deep learning technology and entrepreneurial intention (=0.262, p 0.001); hence, hypotheses 6, 7, and 8 are validated. This provides support for the hypothesized results, claiming that facial recognition plays a significant role in determining the relationship between deep learning technology and entrepreneurial intention. The results would be significant for university students, academicians, and policymakers as they seek to link practices and outcomes. The study is among the pioneers on issues related to deep learning mechanisms in eastern China.

Multi-Sensor Fusion Simultaneous Localization Mapping Based on Deep Reinforcement Learning and Multi-Model Adaptive Estimation.

In this study, we designed a multi-sensor fusion technique based on deep reinforcement learning (DRL) mechanisms and multi-model adaptive estimation (MMAE) for simultaneous localization and mapping (SLAM). The LiDAR-based point-to-line iterative closest point (PLICP) and RGB-D camera-based ORBSLAM2 methods were utilized to estimate the localization of mobile robots. The residual value anomaly detection was combined with the Proximal Policy Optimization (PPO)-based DRL model to accomplish the optimal adjustment of weights among different localization algorithms. Two kinds of indoor simulation environments were established by using the Gazebo simulator to validate the multi-model adaptive estimation localization performance, which is used in this paper. The experimental results of the proposed method in this study confirmed that it can effectively fuse the localization information from multiple sensors and enable mobile robots to obtain higher localization accuracy than the traditional PLICP and ORBSLAM2. It was also found that the proposed method increases the localization stability of mobile robots in complex environments.

Denoising Diffusion Probabilistic Models and Transfer Learning for citrus disease diagnosis.

Plant Disease diagnosis based on deep learning mechanisms has been extensively studied and applied. However, the complex and dynamic agricultural growth environment results in significant variations in the distribution of state samples, and the lack of sufficient real disease databases weakens the information carried by the samples, posing challenges for accurately training models. This paper aims to test the feasibility and effectiveness of Denoising Diffusion Probabilistic Models (DDPM), Swin Transformer model, and Transfer Learning in diagnosing citrus diseases with a small sample. Two training methods are proposed: The Method 1 employs the DDPM to generate synthetic images for data augmentation. The Swin Transformer model is then used for pre-training on the synthetic dataset produced by DDPM, followed by fine-tuning on the original citrus leaf images for disease classification through transfer learning. The Method 2 utilizes the pre-trained Swin Transformer model on the ImageNet dataset and fine-tunes it on the augmented dataset composed of the original and DDPM synthetic images. The test results indicate that Method 1 achieved a validation accuracy of 96.3%, while Method 2 achieved a validation accuracy of 99.8%. Both methods effectively addressed the issue of model overfitting when dealing with a small dataset. Additionally, when compared with VGG16, EfficientNet, ShuffleNet, MobileNetV2, and DenseNet121 in citrus disease classification, the experimental results demonstrate the superiority of the proposed methods over existing approaches to a certain extent.