Deep Learning for Image Generation, Medical Imaging, and Anomaly Detection
This community develops deep learning algorithms to generate synthetic images, analyze medical scans, and detect anomalies in industrial and remote sensing data.
The work centers on generative adversarial networks (GANs) for image synthesis and data augmentation, alongside convolutional neural networks for classification and segmentation. Specific applications include reconstructing computed tomography and magnetic resonance images, segmenting brain tumors, and performing fault diagnosis in mechanical systems. The research also addresses anomaly detection and remote sensing image processing. Methods frequently involve transfer learning and handling data scarcity through synthetic data generation.
The largest share of the community's output is found in gallium research, accounting for 10.7% of all gallium research, and 9,611 papers here. Nitrogen research follows with 6.8% of its total output, and radon research with 14.1%.
The community comprises 45,984 papers, publishing most frequently in arXiv, Lecture Notes in Computer Science, and Ultrasonic Imaging.
Recent work continues to focus on AI applications in clinical medicine, plant disease detection, and explainable AI for biomedical imaging, alongside surveys on generative AI challenges and edge AI deployment.