Deep learning solutions for pneumonia detection: performance comparison of custom and transfer learning models

Published: Oct. 21, 2024

Pneumonia is one of the leading causes illness and death worldwide. In clinical practice, Chest X-ray imaging a common method used to diagnose pneumonia. However, traditional pneumonia diagnosis through analysis requires manual annotation by healthcare professionals which delays treatment. This study aimed investigate compare three different deep learning methodologies for classifying detect disease in patients. These advanced models have potential overcome challenges reliability accessibility diagnostic practices. The evaluated included custom convolutional neural network (CNN), transfer approach as well fine-tuning strategy based on ResNet152V2. were rigorously assessed compared across various metrics, including testing accuracy, loss, precision, F1 score, recall. comparative shows that outperforms other methods terms operational effectiveness, with CNN being next most effective, ranking last. also highlights false negatives can more serious consequences than positives, even without specialized medical knowledge.

Language: Английский

---

Leveraging Artificial Intelligence to Enhance Data Security and Combat Cyber Attacks

Deleted Journal, Journal Year: 2024, Volume and Issue: 5(1), P. 392 - 399

Published: Aug. 29, 2024

This research paper examines the potential of artificial intelligence (AI) in strengthening data security and mitigating growing threat cyber-attacks. As digital threats continue to evolve pose significant risks businesses, organizations, government agencies, individual users, there is an urgent need for more robust adaptive measures. study explores how AI can be leveraged enhance network security, focusing on its applications detection, response automation, predictive analysis. Through a comprehensive literature review analysis current AI-driven solutions, this aims provide insights into effectiveness cybersecurity propose strategies implementation. The findings suggest that has significantly improve measures, offering faster accurate risk assessment, enhanced capabilities. However, challenges related implementation, privacy, human oversight are also addressed. contributes body knowledge provides valuable recommendations organizations seeking strengthen their posture increasingly complex landscape.

Language: Английский

---

Research on Driver Facial Fatigue Detection Based on Yolov8 Model

Published: May 31, 2024

Language: Английский

---

Smart Clothing System for Arrhythmia Detection Based on Digital Twin Technology

Lecture notes in computer science, Journal Year: 2025, Volume and Issue: unknown, P. 71 - 82

Published: Jan. 1, 2025

Language: Английский

---

Attention-based backdoor attacks against natural language processing models

Applied Soft Computing, Journal Year: 2025, Volume and Issue: unknown, P. 112907 - 112907

Published: Feb. 1, 2025

Language: Английский

---

FastTalker: Jointly Generating Speech and Conversational Gestures from Text

Lecture notes in computer science, Journal Year: 2025, Volume and Issue: unknown, P. 177 - 194

Published: Jan. 1, 2025

Language: Английский

---

Unifying emotion-oriented and cause-oriented predictions for emotion-cause pair extraction

Neural Networks, Journal Year: 2024, Volume and Issue: 178, P. 106431 - 106431

Published: June 5, 2024

Language: Английский

---

Research on Driver Facial Fatigue Detection Based on Yolov8 Model

Published: June 12, 2024

In a society where traffic accidents frequently occur, fatigue driving has emerged as grave issue. Fatigue detection technology, especially those based on the YOLOv8 deep learning model, seen extensive research and application an effective preventive measure. This paper discusses in depth methods technologies utilized model to detect driver fatigue, elaborates current status both domestically internationally, systematically introduces processing algorithm principles for various datasets. study aims provide robust technical solution preventing detecting driving, thereby contributing significantly reducing safeguarding lives.

Language: Английский

---

Cross- and Context-Aware Attention Based Spatial-Temporal Graph Convolutional Networks for Human Mobility Prediction

ACM Transactions on Spatial Algorithms and Systems, Journal Year: 2024, Volume and Issue: unknown

Published: June 14, 2024

The COVID-19 pandemic has dramatically transformed human mobility patterns. Therefore, prediction for the “new normal” is crucial to infrastructure redesign, emergency management, and urban planning post pandemic. This paper aims predict people’s number of visits various locations in New York City using COVID data past two years. To quantitatively model impact cases on patterns across period, this develops a CCAAT-GCN ( C ross- ontext- A ttention based Spatial-Temporal G raph onvolutional N etworks). proposed validated SafeGraph from August 2020 April 2022. rich set baselines are performed demonstrate performance our model. Results superior method. Also, attention matrix learned by exhibits strong alignment with situation points interest within geographic region. suggests that effectively captures intricate relationships between case rates developed findings can offer insights into pattern future disruptive events pandemics, so as assist preparedness planners, decision-makers policymakers.

Language: Английский

---

EGGesture: Entropy-Guided Vector Quantized Variational AutoEncoder for Co-Speech Gesture Generation

Published: Oct. 26, 2024

Language: Английский

---