Enhanced hybrid attention deep learning for avocado ripeness classification on resource constrained devices

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Jan. 29, 2025

Attention mechanisms such as the Convolutional Block Module (CBAM) can help emphasize and refine most relevant feature maps color, texture, spots, wrinkle variations for avocado ripeness classification. However, CBAM lacks global context awareness, which may prevent it from capturing long-range dependencies or patterns relationships between distant regions in image. Further, more complex neural networks improve model performance but at cost of increasing number layers train parameters, not be suitable resource constrained devices. This paper presents Hybrid Neural Network (HACNN) classifying on It aims to perform local enhancement capture relationships, leading a comprehensive extraction by combining attention modules models. The proposed HACNN combines transfer learning with hybrid mechanisms, including Spatial, Channel, Self-Attention Modules, effectively intricate features fourteen thousand images. Extensive experiments demonstrate that EfficienctNet-B3 significantly outperforms conventional models regarding accuracy 96.18%, 92.64%, 91.25% train, validation, test models, respectively. In addition, this consumed 59.81 MB memory an average inference time 280.67 ms TensorFlow Lite smartphone. Although ShuffleNetV1 (1.0x) consumes least resources, its testing is only 82.89%, insufficient practical applications. Thus, MobileNetV3 Large exciting option has 91.04%, usage 26.52 MB, 86.94 These findings indicated method enhances classification ensures feasibility implementation low-resource environments.

Language: Английский

---

Revolutionizing Pulmonary Diagnostics: A Narrative Review of Artificial Intelligence Applications in Lung Imaging

Cureus, Journal Year: 2024, Volume and Issue: unknown

Published: April 5, 2024

Artificial intelligence (AI) has emerged as a transformative force in healthcare, particularly pulmonary diagnostics. This comprehensive review explores the impact of AI on revolutionizing lung imaging, focusing its applications detecting abnormalities, diagnosing conditions, and predicting disease prognosis. We provide an overview traditional diagnostic methods highlight importance accurate efficient imaging for early intervention improved patient outcomes. Through lens AI, we examine machine learning algorithms, deep techniques, natural language processing analyzing radiology reports. Case studies examples showcase successful implementation diagnostics, alongside challenges faced lessons learned. Finally, discuss future directions, including integrating into clinical workflows, ethical considerations, need further research collaboration this rapidly evolving field. underscores potential enhancing accuracy, efficiency, accessibility healthcare.

Language: Английский

---

Multi-axis transformer based U-Net with class balanced ensemble model for lung disease classification using X-ray images

Journal of X-Ray Science and Technology, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 26, 2025

Background: Chest X-rays are an essential diagnostic tool for identifying chest disorders because of its high sensitivity in detecting pathological anomalies the lungs. Classification models based on conventional Convolutional Neural Networks (CNNs) adversely affected due to their localization bias. Objective: In this paper, a new Multi-Axis Transformer U-Net with Class Balanced Ensemble (MaxTU-CBE) is proposed improve multi-label classification performance. Methods: This may be first attempt simultaneously integrate benefits hierarchical into encoder and decoder traditional U-shaped structure improving semantic segmentation superiority lung image. Results: A key element MaxTU-CBE Contextual Fusion Engine (CFE), which uses self-attention mechanism efficiently create global interdependence between features various scales. Also, deep CNN incorporate ensemble learning address issue class unbalanced learning. Conclusions: According experimental findings, our suggested outperforms competing BiDLSTM classifier by 1.42% CBIR-CSNN techniques 5.2%

Language: Английский

---

CKE-Former: A clinical knowledge-enhanced transformer for disease classification in telemedicine

Knowledge-Based Systems, Journal Year: 2025, Volume and Issue: unknown, P. 113259 - 113259

Published: March 1, 2025

Language: Английский

---

Bridging efficiency and interpretability: Explainable AI for multi-classification of pulmonary diseases utilizing modified lightweight CNNs

Image and Vision Computing, Journal Year: 2025, Volume and Issue: unknown, P. 105553 - 105553

Published: April 1, 2025

Language: Английский

---

Analysis of Lung Disease Prediction using Machine Learning Algorithms

International Journal of Advanced Research in Science Communication and Technology, Journal Year: 2024, Volume and Issue: unknown, P. 425 - 4234

Published: Oct. 26, 2024

Lung diseases are a notable in global health concern, requiring early diagnosis for better recovery and survival rates. Deep learning strategies, especially CNNs, have shown great promise self lung disease from medical images like chest X-rays. Ensemble methods using pretrained networks such as VGG16, InceptionV3, MobileNetV2 achieved up to 94% accuracy identifying conditions COVID-19, pneumonia, opacity. Lightweight CNN models also performed well, with 89.89%. Traditional machine algorithms, including Random Forest Logistic Regression, yielded rates between 88% 90%. A hybrid deep approach, combining based feature extraction classifiers AdaBoost, SVM, Forest, improved classification by 3.1% reduced computational complexity 16.91%. This method highlights the main integrating traditional enhance detection efficiency

Language: Английский

---

Deep Ensemble Learning Model for Diagnosis of Lung Diseases from Chest X -Ray Images

Indian Journal of Science and Technology, Journal Year: 2024, Volume and Issue: 17(8), P. 702 - 712

Published: Feb. 15, 2024

Objectives: This study aims to develop a robust medical recognition system using deep learning for the identification of various lung diseases, including COVID-19, pneumonia, opacity, and normal states, from chest X-ray images. The focus is on implementing ensemble fixed features methods enhance diagnostic capabilities, contributing development cost-effective reliable tool combating global epidemic disorders. Methods: utilizes Kaggle dataset containing COVID-19 radiography Raw images undergo preprocessing contrast enhancement noise removal while addressing imbalance through near-miss resampling. Ensemble techniques, two three-level methods, are employed harness strengths individual base learners—VGG16, InceptionV3, MobileNetV2. model's performance evaluated metrics such as accuracy, recall, precision, F1-score. For remote access, user interface shared web link developed Python Gradio. Findings: In two-level ensembles, learners concatenated classified support vector machine. Three-level ensembles use by three machine classifiers, employing majority voting final prediction. method achieved 93% F1 score. model demonstrates superior performance, achieving 94% accuracy in detecting four namely states. Novelty: research contributes field showcasing efficacy technology, particularly learning, enhancing detection diseases raw employs modified efficient pretrained networks automatic feature extraction, eliminating need manual engineering. stands promising decision-support healthcare professionals, low-resource environments. Keywords: Convolutional Neural Network (CNN), Deep Learning (DL), Transfer (TL), (EL), Fixed Chest X­rays (CXR), Lung

Language: Английский

---

Improving Early Detection and Classification of Lung Diseases With Innovative MobileNetV2 Framework

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 116202 - 116217

Published: Jan. 1, 2024

Any condition that damage or impedes the normal operation of lungs is classified as a lung disease, and failure to identify address it early on can potentially lead false outcomes. To this challenge, two innovative techniques are proposed for disease classification, supporting medical professionals diagnose provide preventive measures at an stage. The Proposed Model 1 integrates custom MobileNetV2L2 architecture, builds upon MobileNetV2 framework through fine-tuning customization. This model incorporates ridge L2 regularizer within its dense layer enhance performance. 2, CNN2 built CNN foundational block, fine-tuned with ELU activation function, replacing ReLU, regularization technique. research utilizes publicly available datasets: DS1(Data Set1), which Lung Disease 5-class dataset, DS2(Data Set2), 4-class dataset collected from Kaggle. results better performance than state-of-the-art like EfficientNet B0, InceptionV3, ResNet, InceptionResNetV2. It achieved training accuracy 99.53%, validation 100%, test 95.51%. 2 provides outstanding performance, 96.79%, 91.56%, testing reaching 99.26% serves valuable tool Pulmonologists, providing secondary opinion in diagnostic process.

Language: Английский

---

Multi-Label Classification of Lung Diseases Using Deep Learning

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 124062 - 124080

Published: Jan. 1, 2024

Language: Английский

---

Detection of multi‐class lung diseases based on customized neural network

Computational Intelligence, Journal Year: 2024, Volume and Issue: 40(2)

Published: April 1, 2024

Abstract In the medical image processing domain, deep learning methodologies have outstanding performance for disease classification using digital images such as X‐rays, magnetic resonance imaging (MRI), and computerized tomography (CT). However, accurate diagnosis of by personnel can be challenging in certain cases, complexity interpretation non‐availability expert personnel, difficulty at pixel‐level analysis, etc. Computer‐aided diagnostic (CAD) systems with proper training shown potential to enhance accuracy efficiency. With exponential growth data, CAD analyze extract valuable information assisting during process. To overcome these challenges, this research introduces CX‐RaysNet, a novel deep‐learning framework designed automatic identification various lung classes chest X‐ray images. The core novelty CX‐RaysNet lies integration both convolutional group layers, along usage small filter sizes incorporation dropout regularization. This phenomenon helps us optimize model's ability distinguish minute features that reveal different diseases. Additionally, data augmentation techniques are implemented augment testing datasets, which enhances robustness generalizability. evaluation reveals promising results, proposed model achieving multi‐class 97.25%. Particularly, study represents first attempt specifically low‐power embedded devices, aiming improve detection while minimizing computational resources.

Language: Английский

---