Deep learning can improve non-invasive alopecia diagnosis using hair images.
Transfer learning with three neural network architectures accurately classifies hair diseases.
The method creates realistic, anonymous acne face images for research, achieving 97.6% accuracy in classification.
4 citations
,
April 2024 in “Complex & Intelligent Systems” NLKFill improves high-resolution image inpainting by effectively capturing image details and enhancing speed.
November 2022 in “bioRxiv (Cold Spring Harbor Laboratory)” Using deep learning to predict gene expression from images could help assess colorectal cancer metastasis.